mirror of
https://github.com/Akkudoktor-EOS/EOS.git
synced 2026-11-13 09:13:30 +01:00
7e10f64048657795271df2a62b1cfbb92ca0329c
84
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 Bobby Noelte
|
7e10f64048 |
chore: development version update (incl. home assistant config)
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
 Bobby NoelteandClaude Opus 5.5
|
88904874e7 |
fix(version): keep VERSION_DATE in _version_date.py up to date
Additional fixes Use a portable shebang instead of the workspace .venv in update_version.py Update the stale VERSION_DATE of 2026-02-24 to 2026-11-12 Assisted-by: Claude Code (claude-opus-5-5) Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
Bobby NoelteandClaude Opus 5.5
|
e255f987bf |
feat(devices): add battery profiles
A battery profile describes a battery model: the charge and discharge limit
maps over temperature and SoC, the charge and discharge efficiency curves
over C-rate, the technical SoC bounds, the hard temperature cutoffs and,
for a model, its nominal capacity and power limits. EOS packages the
generic profile generic-lfp; users add their own profiles.
A battery or electric vehicle selects its profile by battery_profile and
describes how the installation differs from it in profile_adjustment
(capacity, power limits of the installation, periodic full charge). The
effective values are resolved from both at the measured operating point
(temperature, SoC, power) and shown as effective_* fields.
The GENETIC simulation uses the limits and efficiencies of the profile over
SoC, with precomputed tables as a fast path; GENETIC0 gets the values at the
operating point. Optimization parameters are always valid: a battery whose
profile is missing, unreadable or invalid uses the packaged generic-lfp
profile, and the former defaults only if that is not usable either; each is
logged as error, and an optimization run is never cancelled because of a
battery.
Additional changes
- Add a generic device profile store: user and packaged profiles, atomic
saves, compiled profiles cached per file and checked again at most once a
second, plain file names as profile names ('schema' is reserved). On
Windows, a save or read that collides with the other is retried briefly.
- Add the REST API /v1/profile to list the profile kinds and to list, read
(document and JSON schema), save and delete profiles.
- Check the hard temperature cutoffs on the exact temperature.
- Resolve inhibited charging or discharging to a power of 0, e.g. for a cold
battery; give GENETIC0 its own discharge limit for that.
- Limit generic profiles without own power limits to the former 5000 W.
- Accept only the implemented units; current limits need the nominal
voltage.
- Migrate legacy battery settings into the profile adjustment, keep the
former charge power limit as discharge limit (also for a Home Assistant
entity on it) and report dropped efficiencies as errors.
- Limit GENETIC battery discharge by the discharge power.
- Allow battery profile lookups while the configuration is created.
- Resolve a battery once per configuration dump.
- Package the bundled profiles, document battery profiles and regenerate
the configuration and OpenAPI documentation.
- Ignore the report PDFs written by the visualization tests.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Assisted-by: Claude Code (claude-opus-5-5)
|
||
|
Bobby NoelteandClaude Opus 5.5
|
2c58874c52 |
fix(cache): use a reproducible function hash for cache_in_file keys
Build the cache_in_file key from an MD5 hash of the wrapped function instead of inspect.getsource(). Reading the source fails for functions whose source is not on disk, and the key could go stale when the file changes after import. The hash covers what decides the result besides the arguments: the qualified name, the bytecode with the names of called functions, attributes and globals, the constants including nested functions, the defaults and the values captured from an enclosing function. Wrappers made with functools.wraps are unwrapped. Code objects, functions and sets are hashed without memory addresses or hash-seed dependent order, so the key stays the same across restarts and the file cache keeps hitting. Additional fixes - Build the key from the arguments bound to the signature, so *args, **kwargs and the arguments of wrapped functions are part of it. - Apply control parameters given at call time to that call only; they no longer change the decorator settings for later calls. - Apply all control parameters of a call, so until_date is no longer dropped when force_update is also passed; keep the precedence until_datetime > with_ttl > until_date. - Recreate an unreadable cache file in that call only, instead of disabling the cache for the rest of the process. - Pass force_update in two provider tests that relied on a force_update left over from an earlier test. - Fix "Calback" typo in cache_energy_management_store_callback docstring. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
Bobby NoelteandClaude Opus 5.5
|
57132df44e |
fix(config): enable documentation mode only explicitly
ConfigEOS.documentation_mode() was true as soon as Sphinx was imported anywhere in the process. A test importing Sphinx therefore switched all later tests into documentation mode, and test_config_md_current failed depending on test order. Documentation mode is now enabled only by the documentation generators (_force_documentation_mode) or by the Sphinx configuration, which sets sys._called_from_sphinx in docs/conf.py. Declare _force_documentation_mode as a ClassVar. Without the annotation pydantic turned it into a private attribute, and the class attribute was a truthy descriptor object. Behavior change: because of that descriptor, every running EOS was in documentation mode, so the get_adapter/config/measurement/prediction/ems/ database/resource_registry() guards never raised "access before init" and created the singletons lazily on first access. They now raise for every access before initialization. The worker process of uvicorn's auto-reload (server.reload, make run-dev) relied on that: the app's lifespan now sets up the configuration and initializes the singletons that do not exist yet. Leave runtime state out of the documentation: the Home Assistant entity ID properties and OptimizationCommonSettings.keys return an empty list in documentation mode, so entity IDs and solution keys of a running system no longer reach the generated configuration docs. Keep the tests isolated: test_doc sets documentation mode on the class and test_docstringrst restores sys._called_from_sphinx after reading the Sphinx configuration. Add tests for how documentation mode is enabled and for the runtime values it leaves out. Additional changes - Generate openapi.json in documentation mode, as the config docs. - Share the documentation mode check as coreabc.documentation_mode(). - Move the logging and server address setup of run_eos() to setup_eos_config(). - Test that EOS starts with server.reload. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> |
||
     
|
f804d48072 |
feat(pvforecast): add adaptive per-plane PV learning and diagnostics (#1422)
PV arrays with different orientations and curtailment need separate plant calibration and weather correction. This opt-in provider publishes normal EOS AC/DC forecast keys at 15-minute resolution. - Preserve daily per-plane energy/shape calibration and validated short-term local/trend selection. Curtailment-sensitive arrays require a fresh release flag or an explicitly configured learning guard. - Add switchable empirical solar-angle response relative to separate eligible reference arrays, removing their common weather error. This learns angular residuals, not physical angles or a measured geometric horizon. - Archive actually received weather model inputs, first local receipt and frozen per-member PV candidates. Returned past weather is excluded; unknown source issuance remains explicit. Learn model weights and forecast-regime bias separately at 2–6, 6–24 and 24–48 hours. - New long-horizon candidates remain in shadow until their genuinely issued outputs improve equal-day MAE by the configured margin, win on at least 60% of validation days and pass a matched-support energy-bias guard. Insufficient evidence retains calibration. Daily yields still require complete local days, including DST. - Expose fitted parameters, independent training/validation day counts, errors, selected methods and persistence in the existing status API, with additional normal prediction keys for Grafana. Retain physical DC through angular/weather corrections an enforce individual inverter AC limits. Co-authored-by: Normann <github@koldrack.com> Co-authored-by: Christin <chrissi@von-perbandt.de> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: YMT <yn276364@dal.ca> Co-authored-by: sarankumar1325 <120080287+sarankumar1325@users.noreply.github.com> |
||
|
Bobby NoelteandGitHub
|
7554e2dc47 |
fix(database): make delete_by_datetime(None, None) a true full wipe (#1476)
delete_by_datetime() with no bounds was documented as "delete all
records", but db_delete_records() assumed that
_db_ensure_loaded(start=None, end=None) loads every record into memory
before building the delete list. That is wrong: _db_ensure_loaded()
implements a NONE -> INITIAL -> FULL phase machine, and with both bounds
None in the NONE phase it loads only an *initial window* and leaves the
phase at INITIAL. Records outside that window never enter
_db_sorted_timestamps, so they are never marked for deletion and survive
in the physical store. Symptom: after a "wipe", db_count_records() still
returned N records (observed: 4 leftovers from a previous dense fixture).
A second delete_by_datetime() call happened to escalate INITIAL -> FULL
and deleted the rest, which is why the bug looked intermittent and why
stacked test runs accumulated rows.
Fix: in db_delete_records(), when both bounds are None, call
self._db_load_full() directly; otherwise keep the existing
_db_ensure_loaded(start, end) path. The delete list is then complete and
the "wipe" fixture actually wipes.
delete_by_datetime() remains a soft delete (rows are only added to
_db_deleted_timestamps; the physical DELETE happens on the next flush),
unchanged by this commit.
Besides this an additional fix was added:
* fix(tests): assert UTC-instant ordering in test_timestamps_remain_sorted
The test asserted `dts == sorted(dts)` on seq.records[].date_time, which
are tz-aware DateTime values in Europe/Berlin. Compaction buckets on a
UTC grid but records are materialized in local time, so across a DST
fall-back the local wall-clock sequence is non-monotonic even though the
UTC instants are ordered (e.g. 00:45 UTC -> 02:45 CEST, 01:00 UTC ->
02:00 CET). sorted() compares wall-clock fields, so it flags the pair at
the seam (index 1248 on 2026-10-25 for Europe/Berlin) as out of order.
The assertion is unsatisfiable for any 15-min grid spanning a fall-back
boundary when records are labelled in local time.
The actual invariant is "ascending by UTC instant", which is what
_db_sorted_timestamps guarantees and what the adjacent alignment tests
already assert via int(rec.date_time.timestamp()). Change the assertion
to compare integer UTC epochs:
epochs = [int(r.date_time.timestamp()) for r in seq.records if r.date_time]
assert epochs == sorted(epochs)
Note: this test was seasonally flaky. It only fails while a DST
transition lies inside the [now - 4 weeks, now] window used by
dense_energy_seq, which is why it newly started failing without any code
change. Upstream should ideally emit compacted records in UTC (or order
by UTC instant) so that wall-clock ordering is not silently assumed.
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com>
|
||
|
YTM-21andGitHub
|
8f4fa6fe94 |
fix(config): report computed fields as read-only on update (#1454)
PUT /v1/config/{path} answered "does not exist in model" when the
target was a computed field such as eos_device_instruction_entity_ids.
Raise an explicit read-only error instead, so API clients can tell
read-only fields from unknown ones.
Closes #913
|
||
 
|
7dc1ee92fa |
perf(genetic): faster runs with unchanged results (#1385)
* feat(devices): add load-dependent DC-to-AC efficiency curve for GENETIC inverters Real inverters are much less efficient at low load than at rated power (e.g. 0.80 at 2 % vs. 0.93 at 7 % load). A constant dc_to_ac_efficiency overrates night-time battery discharge into a small base load and, with it, arbitrage and grid charging. Add two optional inverter settings for the GENETIC optimizer: - dc_to_ac_efficiency_curve: list of (load_fraction, efficiency) points, load fraction = AC power / max_power_w. Points must be strictly increasing in [0, 1] with efficiencies in (0, 1]; at least two points. Linear interpolation, clamped at both ends. When set, the curve replaces dc_to_ac_efficiency. - dc_to_ac_efficiency_reference_load_fraction (default 0.06): load at which the curve values stored energy without a specific conversion. Battery discharges in Inverter.process_energy() use the efficiency at the AC energy of each conversion. The end-of-horizon value, the terminal and tail value curves and the AC charge break-even penalty use Inverter.reference_dc_to_ac_efficiency. Without a curve, dc_to_ac_efficiency_at() and the reference efficiency return dc_to_ac_efficiency itself, so results are bit-identical to the previous model. GENETIC0 is unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(genetic): make the fitness cache size configurable The per-run fitness cache of the GENETIC optimizer is unbounded. It grows with individuals x generations (~110k keys at 300 x 400, about 1.1 kB each, 1.6 kB with an EV) and adds 120-170 MB to the peak memory of a run, while only 6-15 % of the lookups hit. On 512 MB boards this is the largest part of the optimizer's memory. Add optimization.genetic.fitness_cache_max_entries: - None (default): unbounded, the previous behaviour. - 0: no cache; every genome is evaluated. - N > 0: at most N keys. The cache is then an OrderedDict and evicts the oldest key first (FIFO, O(1)); LRU would cost bookkeeping on every hit for a cache that rarely hits. A cache hit returns exactly what the deterministic evaluation would return, so the plan for a fixed seed is identical with any setting. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * perf: import heavy optional dependencies on first use A GENETIC server with external (import) providers and EOSdash disabled loads several large packages at start that it never uses, which costs a noticeable part of the memory on 512 MB boards: - matplotlib via the GENETIC/GENETIC0 PDF report modules (server/eos.py) - statsmodels via the Holt-Winters fallback in prediction/priceabc.py - pvlib via the weather base class and the pvlib-based providers - bs4/lxml via the ClearOutside provider - scipy via the self-consumption interpolator singletons, which were loaded at import time - the GENETIC0 optimizer, imported by core/ems.py - tzfpy, whose polygon data the computed general.timezone loads Import these where they are used. The interpolator tables are loaded on first use (thread-safe). Nothing changes in what is computed. general.timezone is computed from latitude/longitude and cannot be set. Add general.timezone_override (IANA name, validated with pendulum): when set, general.timezone returns it and the lookup is skipped. When unset (default) the lookup is unchanged. Tests that patched the moved names now patch them at their source. A new test starts the server module in a subprocess and checks that none of these modules is loaded. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * perf(prediction): evaluate self-consumption tables without SciPy The GENETIC inverter (and the GENETIC0 one) evaluates a pickled scipy.interpolate.RegularGridInterpolator for the self-consumption probabilities. SciPy is used nowhere else in EOS, yet loading it costs about 35 MB RSS (21 MB anonymous) on Linux arm64, which every optimization run pays. Add utils/gridinterpolator.py: - LinearGridInterpolator: 2-D linear interpolation with the same steps as SciPy's linear 2-D path (interval search, normalised distances, corner weights summed in the same order, fill value outside the grid, NaN for NaN input). - load_grid_interpolator(): reads the shipped pickles with a restricted unpickler that maps the SciPy class to a stand-in and accepts only the NumPy globals needed for the arrays. SciPy is not imported, and the data files no longer depend on SciPy's private module paths. Results are bit-identical to SciPy on Linux (aarch64, checked on 3.6M points per table including the query pattern of the inverter). The macOS arm64 SciPy wheel fuses multiply-add, so there SciPy itself differs by <= 1.1e-16; GENETIC plans for fixed seeds are unchanged on both platforms. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(prediction): make the SciPy-free table evaluation opt-in Add optimization.self_consumption_interpolator (scipy | numpy, default scipy). The default keeps evaluating the self-consumption tables with SciPy's RegularGridInterpolator, built from the same grid, values and fill value the pickle holds, so results are unchanged. With numpy the tables are evaluated by LinearGridInterpolator and SciPy is never imported. The backend is read on every evaluation, so a config change applies to the next run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * perf(genetic): cache the load distribution, clone genomes directly Two hot spots of a GENETIC run, found by profiling a 15 minute run with EV and two appliances. The solution is unchanged: both repository reference inputs give a byte-identical GeneticSolution at 400 generations. - SelfConsumptionProbabilityInterpolator._load_distribution depends on the mean load only, but ran the table interpolation over all minute-load bins for every new (load, PV) pair - 27 % of the run. It is now kept per (backend, bounded mean load); a run asks for few loads (one per slot, plus appliance combinations) with ever-changing PV values. The table is static, so entries never go stale; the cache is bounded. Scalar bounds use min/max instead of np.clip. - _make_offspring cloned parents with copy.deepcopy (8 % of the run), although the child's fitness, extra_data and immigrant protection are dropped right after. A genome is a flat list of ints, so the child is created from the parent's genes directly. Measured (1 process, Apple M5): eHive run with EV + 2 appliances, 200 generations 33.7 s -> 25.1 s; EV only 34.1 s -> 24.3 s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): cheaper slot simulation, lean lookup cache, batch duplicates Four hot spots of a GENETIC run, found by profiling a 15 minute run with EV and two appliances. The solution is unchanged: both repository reference inputs (optimize_input_1/2, seed 42, 400 generations) give a byte-identical solution before and after, as do the fixed-seed runs with 1, 2 and 3 evaluation processes. * Expected direct consumption: the lookup went through the shared cache_energy_management store on every slot of every candidate - a key built from the callable, a callback per call and room for the last 100 results only (18 % misses). It is now a dictionary of the interpolator, dropped together with that store at the start of an energy management run (CacheEnergyManagementStore.generation), bounded to 8192 entries. * Slot simulation: per-slot inputs are read as plain Python numbers and per-slot results are collected in lists, turned into arrays after the loop. Element access on NumPy arrays and NumPy scalar arithmetic were the largest share of the loop. process_energy() skips the charge/export block when a slot has no PV surplus and bounds values without min()/max() calls (exact equivalents, NaN and signed zero included). * Genome keys: bytes() does the 0-255 range check in C, and a genome that is relevant as a whole is packed without copying it first. * Identical genomes within a generation are evaluated once also when the fitness cache is off (fitness_cache_max_entries = 0), by grouping the batch. It needs no memory beyond the batch; the cache costs 85-95 MB. Same run, one process, 60 generations (Apple M-series): 7.8 s -> 5.9 s with appliances, 7.1 s -> 5.0 s without. 200 generations with 3 evaluation processes: 11.4 s -> 8.2 s. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(devices): keep the battery charge bookkeeping in plain numbers Battery.charge_energy() runs for every slot of every candidate plan. The per-slot charge limit and charged energy were NumPy arrays read and written one element at a time, and the NumPy scalars they return made the state of charge arithmetic slow from then on. They are plain lists now. The lower charge rates tried when a request does not fit are taken from a tuple built once per charge_rates array instead of a searchsorted() plus generator per request. The results are unchanged (same IEEE doubles, same order of operations). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(devices): check the discharge release first, plain discharge bookkeeping Inverter._discharge_battery_to_ac() looked up the conversion efficiency before the battery told it that the plan does not release discharge in this slot. Battery.discharge_released() answers that first; the result of such a slot was (0.0, 0.0) before and is now returned directly. The discharged energy per slot is a plain list instead of a NumPy array read and written one element at a time. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(devices): find the efficiency curve segment by bisection Inverter.dc_to_ac_efficiency_at() runs for every battery discharge of every candidate plan and walked the curve points one by one. The load fractions are split off once per curve and the segment is found by bisection; the segment and the interpolation are those of interpolate_efficiency_curve(). Compared bit for bit against that function for 200000 load values, the curve points themselves, zero, infinities and NaN. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): sum appliance loads once per candidate, read the SoC directly The slot loop asked every appliance for its load in every slot (one method call with a range check each) and converted the state of charge to percent through a method call twice per slot. The aggregate appliance load is now summed once per candidate, in the order the loop added the appliances, and the state of charge is read with the Battery formula in place. Appliances of another class or with a curve that does not cover the simulated slots, and batteries with their own current_soc_percentage(), keep the calls. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): compute the control end slot once per run control_end_slot is a property that reads the interval and the horizon from the configuration on every access - several hundred thousand times per run. The configuration is frozen for the duration of a run, so the value is computed once when the snapshot is taken and dropped with it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): work out the energy-shift targets once per source slot _energy_shift_target_slots() runs for every export slot of every energy-shift mutation, in the server process while the evaluation workers wait. Which later slots are worth the shift, and in which order, depends on the forecasts and the source slot only; it is now worked out once per source slot and run, and the individual only filters the ordered slots. Filtering the sorted superset gives the same list as sorting the filtered slots (the sort is stable and its key does not depend on the individual), so mutations and the random number stream are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(devices): compute the battery discharge of a slot in one go Serving load or exporting from the battery went through three nested calls per slot: Inverter._discharge_battery_to_ac() -> dc_to_ac_efficiency_at() -> Battery.discharge_energy(), with the release of the slot checked twice and the efficiency function called also without a curve. For the plain Battery the inverter now computes the release check, the efficiency and the discharge in one go; other battery implementations keep the step-by-step path. The arithmetic is that of Battery.discharge_energy(). The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): decode appliance starts once per gene combination Every candidate decoded its appliance start genes (feasibility repair over cycle windows and minimum gaps) and rebuilt each appliance's load curve over the whole horizon. The decoded starts, the canonical gene values and the curves depend on the appliance genes only, and a run sees few different combinations among tens of thousands of candidates. Each combination is now decoded and built once per layout (bounded to 4096 combinations) and looked up afterwards. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): compute only the terminal credit per candidate Every evaluation called _terminal_value(), which returns the credit for the energy left in the battery together with a TerminalValueResult report (component values, diagnostics, curves). The evaluation used the credit and dropped the report. _terminal_value_credit() computes the same credit without building the report; the report is still built for the final solution. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): decode battery states through a lookup table decode_charge_discharge() decodes every slot on its own from its state, with about twenty array operations per candidate. The four results per state are now worked out once per state layout and looked up for the slots of a candidate. States outside the table keep the explicit decoding. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): check for a full EV before splitting the genome _repair_ev_charge_at_full_soc() split every candidate's genome and walked all slots in Python to find EV charging in slots that begin at full SoC. Nearly no candidate has such a slot. The full slots are now found on the simulated SoC array first; the genome is split and only those slots are looked at when there are any. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * feat(genetic): evaluate candidates in parallel worker processes A GENETIC run evaluated every candidate in the server process, so a run used one CPU core however many the machine had. On small ARM boards (4x Cortex-A55, thermally limited) a 300x200 run took 12-15 minutes. Candidates are now evaluated in up to two worker processes, forked once per run after the run is fully prepared. Each worker owns a copy of the prepared optimizer; only genomes and results cross the process boundary. The fitness cache stays in the server process: hits are answered there, and identical genomes in a batch are evaluated once, as a serial run would hit the cache for the repeats. Results are applied only after the whole batch returned. If a batch fails or times out, the run continues serially. For a fixed seed the result is identical to a serial run. optimization.genetic.workers sets the number of processes. Unset (default) is automatic: one CPU core always stays free and at most two are used (1-2 cores -> 1, 3 or more -> 2), capped by a cgroup v2 CPU limit of the process or any parent (container limit, systemd CPUQuota). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * feat(ems): notify a URL after every completed optimization A client that consumes EOS plans (here DVhub) had to poll /v1/energy-management/optimization/solution to notice a new plan. That delays the plan by up to the poll interval and makes EOS serialize the complete solution on every poll, also while it is computing. ems.notify_url (default None = off) receives an HTTP POST with a small JSON event ({"event": "optimization_completed", "algorithm", "completed_at", "duration_s"}) when a run has published its solution. The POST runs in a daemon thread with a 5 s timeout, so a slow or unreachable receiver never delays energy management; a failure is logged once per failure streak. Only http(s) URLs are accepted. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(genetic): fewer calls per slot in the simulation loop Three calls ran for every slot of every candidate although most slots do not need them: the AC charge factor was capped by the inverter also when the slot has no AC charging, the battery was asked for its discharged energy through a method, and process_energy() got its export arguments by keyword. Slots without AC charging skip the cap, the discharged energy is read with the Battery formula in place, and the arguments are positional. reset() of a battery creates its two control arrays with zeros() instead of full(). The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(genetic): freeze the optimization configuration for the duration of a run The genetic optimization reads horizon_hours, tail_horizon_hours, interval_sec and prediction.hours from the live configuration on every access (control_slots, tail_slots, ... are properties). A client that updates the configuration while a run is in progress - DVhub shrinks or grows the control horizon as the known price horizon moves during the night - changed these values mid-run. Genomes and device arrays sized at the start then no longer matched the simulation loop and the final evaluation failed: IndexError: index 84 is out of bounds for axis 0 with size 84 Long runs on slow hardware hit this almost every night. optimize_ems() now takes a deep copy of the genetic configuration and the prediction hours at the start of a run, all reads in genetic.py go through it, and it is released when the run ends (also on errors). New values apply to the next run. Tests: tests/test_genetic_config_frozen_per_run.py reproduces the error without the fix (horizon 20 -> 24 mid-run: index 20 out of bounds) and passes with it; genetic/optimization tests 439 passed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(genetic): pin evaluation workers, keep the server responsive Workers are pinned to the last `workers` CPU cores (new setting optimization.genetic.pin_workers, default on); the server process stays unpinned. The run never spreads beyond `workers` cores without a CPU quota, and the server and the rest of the system keep the other cores. No pinning without a spare core. With a CPU limit (container cpu.max, systemd CPUQuota) the automatic worker count now keeps a quarter core of it for the server process (limit 2 -> 1 worker, 2.5 -> 2): two workers used a 2-core limit fully. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(ems): tolerate settings without notify_url Test doubles and older configuration objects have no ems.notify_url. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * perf(devices): bound the PV surplus and export values without min()/max() The rest of Inverter.process_energy() - PV surplus, curtailment and battery export - bounded its values with min()/max() calls. They are conditional statements now, as the first half of the method already is: 'if x < 0.0: x = 0.0' is max(x, 0.0) and 'b if b < a else a' is min(a, b), NaN and signed zero included. The results are unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(devices): evaluate the efficiency curve at the delivered AC load When the battery limits a discharge (low state of charge or its power limit), the conversion ran with the efficiency of the requested AC energy. With a 1000 Wh battery at 10 % SoC, a 1000 W inverter and the curve [(0, .8), (.5, .95), (1, .95)], a request of 500 Wh AC returned 95 Wh AC from the available 100 Wh DC, although the curve gives 82.5 % at that load. The delivered AC energy now solves a = dc * efficiency(a / rated) exactly on the piecewise-linear curve (solve_ac_for_dc_energy), bounded by the request. Unlimited discharges and inverters without a curve are unchanged. Regression tests use a real Battery: low SoC and a power-limited discharge, each with 60- and 15-minute slots, battery discharge losses, the load path through process_energy, and constant efficiency. * docs: regenerate OpenAPI for the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix --------- Co-authored-by: Christin <christin@BigBrother.bikini.bottom.zone> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Andreas <drbacke@gmx.de> |
||
|
|
70b72bfbce |
feat(ems): notify a URL after every completed optimization (#1383)
* feat(devices): add load-dependent DC-to-AC efficiency curve for GENETIC inverters Real inverters are much less efficient at low load than at rated power (e.g. 0.80 at 2 % vs. 0.93 at 7 % load). A constant dc_to_ac_efficiency overrates night-time battery discharge into a small base load and, with it, arbitrage and grid charging. Add two optional inverter settings for the GENETIC optimizer: - dc_to_ac_efficiency_curve: list of (load_fraction, efficiency) points, load fraction = AC power / max_power_w. Points must be strictly increasing in [0, 1] with efficiencies in (0, 1]; at least two points. Linear interpolation, clamped at both ends. When set, the curve replaces dc_to_ac_efficiency. - dc_to_ac_efficiency_reference_load_fraction (default 0.06): load at which the curve values stored energy without a specific conversion. Battery discharges in Inverter.process_energy() use the efficiency at the AC energy of each conversion. The end-of-horizon value, the terminal and tail value curves and the AC charge break-even penalty use Inverter.reference_dc_to_ac_efficiency. Without a curve, dc_to_ac_efficiency_at() and the reference efficiency return dc_to_ac_efficiency itself, so results are bit-identical to the previous model. GENETIC0 is unchanged. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(genetic): make the fitness cache size configurable The per-run fitness cache of the GENETIC optimizer is unbounded. It grows with individuals x generations (~110k keys at 300 x 400, about 1.1 kB each, 1.6 kB with an EV) and adds 120-170 MB to the peak memory of a run, while only 6-15 % of the lookups hit. On 512 MB boards this is the largest part of the optimizer's memory. Add optimization.genetic.fitness_cache_max_entries: - None (default): unbounded, the previous behaviour. - 0: no cache; every genome is evaluated. - N > 0: at most N keys. The cache is then an OrderedDict and evicts the oldest key first (FIFO, O(1)); LRU would cost bookkeeping on every hit for a cache that rarely hits. A cache hit returns exactly what the deterministic evaluation would return, so the plan for a fixed seed is identical with any setting. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * perf: import heavy optional dependencies on first use A GENETIC server with external (import) providers and EOSdash disabled loads several large packages at start that it never uses, which costs a noticeable part of the memory on 512 MB boards: - matplotlib via the GENETIC/GENETIC0 PDF report modules (server/eos.py) - statsmodels via the Holt-Winters fallback in prediction/priceabc.py - pvlib via the weather base class and the pvlib-based providers - bs4/lxml via the ClearOutside provider - scipy via the self-consumption interpolator singletons, which were loaded at import time - the GENETIC0 optimizer, imported by core/ems.py - tzfpy, whose polygon data the computed general.timezone loads Import these where they are used. The interpolator tables are loaded on first use (thread-safe). Nothing changes in what is computed. general.timezone is computed from latitude/longitude and cannot be set. Add general.timezone_override (IANA name, validated with pendulum): when set, general.timezone returns it and the lookup is skipped. When unset (default) the lookup is unchanged. Tests that patched the moved names now patch them at their source. A new test starts the server module in a subprocess and checks that none of these modules is loaded. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * perf(prediction): evaluate self-consumption tables without SciPy The GENETIC inverter (and the GENETIC0 one) evaluates a pickled scipy.interpolate.RegularGridInterpolator for the self-consumption probabilities. SciPy is used nowhere else in EOS, yet loading it costs about 35 MB RSS (21 MB anonymous) on Linux arm64, which every optimization run pays. Add utils/gridinterpolator.py: - LinearGridInterpolator: 2-D linear interpolation with the same steps as SciPy's linear 2-D path (interval search, normalised distances, corner weights summed in the same order, fill value outside the grid, NaN for NaN input). - load_grid_interpolator(): reads the shipped pickles with a restricted unpickler that maps the SciPy class to a stand-in and accepts only the NumPy globals needed for the arrays. SciPy is not imported, and the data files no longer depend on SciPy's private module paths. Results are bit-identical to SciPy on Linux (aarch64, checked on 3.6M points per table including the query pattern of the inverter). The macOS arm64 SciPy wheel fuses multiply-add, so there SciPy itself differs by <= 1.1e-16; GENETIC plans for fixed seeds are unchanged on both platforms. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(prediction): make the SciPy-free table evaluation opt-in Add optimization.self_consumption_interpolator (scipy | numpy, default scipy). The default keeps evaluating the self-consumption tables with SciPy's RegularGridInterpolator, built from the same grid, values and fill value the pickle holds, so results are unchanged. With numpy the tables are evaluated by LinearGridInterpolator and SciPy is never imported. The backend is read on every evaluation, so a config change applies to the next run. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(genetic): evaluate candidates in parallel worker processes A GENETIC run evaluated every candidate in the server process, so a run used one CPU core however many the machine had. On small ARM boards (4x Cortex-A55, thermally limited) a 300x200 run took 12-15 minutes. Candidates are now evaluated in up to two worker processes, forked once per run after the run is fully prepared. Each worker owns a copy of the prepared optimizer; only genomes and results cross the process boundary. The fitness cache stays in the server process: hits are answered there, and identical genomes in a batch are evaluated once, as a serial run would hit the cache for the repeats. Results are applied only after the whole batch returned. If a batch fails or times out, the run continues serially. For a fixed seed the result is identical to a serial run. optimization.genetic.workers sets the number of processes. Unset (default) is automatic: one CPU core always stays free and at most two are used (1-2 cores -> 1, 3 or more -> 2), capped by a cgroup v2 CPU limit of the process or any parent (container limit, systemd CPUQuota). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * feat(ems): notify a URL after every completed optimization A client that consumes EOS plans (here DVhub) had to poll /v1/energy-management/optimization/solution to notice a new plan. That delays the plan by up to the poll interval and makes EOS serialize the complete solution on every poll, also while it is computing. ems.notify_url (default None = off) receives an HTTP POST with a small JSON event ({"event": "optimization_completed", "algorithm", "completed_at", "duration_s"}) when a run has published its solution. The POST runs in a daemon thread with a 5 s timeout, so a slow or unreachable receiver never delays energy management; a failure is logged once per failure streak. Only http(s) URLs are accepted. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(genetic): freeze the optimization configuration for the duration of a run The genetic optimization reads horizon_hours, tail_horizon_hours, interval_sec and prediction.hours from the live configuration on every access (control_slots, tail_slots, ... are properties). A client that updates the configuration while a run is in progress - DVhub shrinks or grows the control horizon as the known price horizon moves during the night - changed these values mid-run. Genomes and device arrays sized at the start then no longer matched the simulation loop and the final evaluation failed: IndexError: index 84 is out of bounds for axis 0 with size 84 Long runs on slow hardware hit this almost every night. optimize_ems() now takes a deep copy of the genetic configuration and the prediction hours at the start of a run, all reads in genetic.py go through it, and it is released when the run ends (also on errors). New values apply to the next run. Tests: tests/test_genetic_config_frozen_per_run.py reproduces the error without the fix (horizon 20 -> 24 mid-run: index 20 out of bounds) and passes with it; genetic/optimization tests 439 passed. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> * feat(genetic): pin evaluation workers, keep the server responsive Workers are pinned to the last `workers` CPU cores (new setting optimization.genetic.pin_workers, default on); the server process stays unpinned. The run never spreads beyond `workers` cores without a CPU quota, and the server and the rest of the system keep the other cores. No pinning without a spare core. With a CPU limit (container cpu.max, systemd CPUQuota) the automatic worker count now keeps a quarter core of it for the server process (limit 2 -> 1 worker, 2.5 -> 2): two workers used a 2-core limit fully. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(ems): tolerate settings without notify_url Test doubles and older configuration objects have no ems.notify_url. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01EYWR1HVM8a8src23xuyL3T * fix(devices): evaluate the efficiency curve at the delivered AC load When the battery limits a discharge (low state of charge or its power limit), the conversion ran with the efficiency of the requested AC energy. With a 1000 Wh battery at 10 % SoC, a 1000 W inverter and the curve [(0, .8), (.5, .95), (1, .95)], a request of 500 Wh AC returned 95 Wh AC from the available 100 Wh DC, although the curve gives 82.5 % at that load. The delivered AC energy now solves a = dc * efficiency(a / rated) exactly on the piecewise-linear curve (solve_ac_for_dc_energy), bounded by the request. Unlimited discharges and inverters without a curve are unchanged. Regression tests use a real Battery: low SoC and a power-limited discharge, each with 60- and 15-minute slots, battery discharge losses, the load path through process_energy, and constant efficiency. * docs: regenerate OpenAPI for the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * docs: regenerate OpenAPI after merging the efficiency-curve fix * test(genetic): cover serial fallback on platforms without fork --------- Co-authored-by: Christin <christin@BigBrother.bikini.bottom.zone> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> Co-authored-by: Andreas <drbacke@gmx.de> |
||
|
|
3651d33915 |
perf: import heavy optional dependencies on first use (#1377)
A GENETIC server with external (import) providers and EOSdash disabled loads several large packages at start that it never uses, which costs a noticeable part of the memory on 512 MB boards: - matplotlib via the GENETIC/GENETIC0 PDF report modules (server/eos.py) - statsmodels via the Holt-Winters fallback in prediction/priceabc.py - pvlib via the weather base class and the pvlib-based providers - bs4/lxml via the ClearOutside provider - scipy via the self-consumption interpolator singletons, which were loaded at import time - the GENETIC0 optimizer, imported by core/ems.py - tzfpy, whose polygon data the computed general.timezone loads Import these where they are used. The interpolator tables are loaded on first use (thread-safe). Nothing changes in what is computed. general.timezone is computed from latitude/longitude and cannot be set. Add general.timezone_override (IANA name, validated with pendulum): when set, general.timezone returns it and the lookup is skipped. When unset (default) the lookup is unchanged. Tests that patched the moved names now patch them at their source. A new test starts the server module in a subprocess and checks that none of these modules is loaded. Co-authored-by: Christin <christin@BigBrother.bikini.bottom.zone> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com> |
||
|
NormannandGitHub
|
fca07a0c62 |
fix: show Home Assistant add-on version in EOSdash (#1356)
Align add-on and EOSdash version. |
||
 
|
04f28997ea |
feat: deliver complete GENETIC optimization to main (#1330)
* feat: adapt configuration for multi optimization algorithms Decouple configuration from optimization algorithm parameters. Add to_[algorithm]_param() methods to the configuration that derive optimization algorithm specific parameters from the configuration. Add x-scope tags to the configuration options that describe for which specific algorithms the configuration option is for. The whole device settings are restructured. There are now general settings for the device classes with the afore mentioned to_[algorithm]_param() methods. The general device settings got their own directory `devices/settings`. By this the parameter class also does not have to be a pydantic model which can be used for future optimization/ simulations speed up. Also the parameter class for a device is now part of the device module. This better decouples and also is the natural place for parameters of a device. Besides this feature there are also fixes and improvements: * feat: extend home appliance time window settings and simulation Home appliance can now be configured for multiple runs with per-cycle allowed time windows. The number of remaining cycles to plan is determined at runtime by reading the ``cycles_completed_measurement_key`` from the measurement store. * feat: specialiced CycleTimeWindowSequence for time window sequences Sequence of time windows associated to cycles. This model specializes ``ValueTimeWindowSequence`` so that the ``value`` field of each ``ValueTimeWindow`` encodes the **cycle index** (0-based integer) the window belongs to. Typical use: an appliance that must run ``n`` times per day, each run constrained to a distinct time window. Assign ``value=0`` to windows for the first cycle, ``value=1`` for the second, and so on. Multiple windows may share the same cycle index (their allowed regions are unioned). Windows with ``value=None`` are silently ignored by all cycle-aware methods. * fix: Make test_configmigrate also regard the _ANY_SENTENIEL in key values * chore: Make devices configurations a map instead of a list This makes config paths stable regardless of declaration order and lets each device settings class build its own config path from ``self.device_id`` without needing an external index. Tests are adapted likewise. Devices configurations are automatically migrated from lists to maps. * chore: rename levelized_cost_of_storage_kwh to levelized_cost_of_storage_amt kwh This better fits in the naming scheme and also makes clear the costs are money. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> * fix: runtime config update ignored by config file Runtime settings were handed back to pydantic-settings as init settings, which rank below the config file and the environment. Any key already present in EOS.config.json or in the environment silently discarded the update, so a bulk PUT /v1/config returned 200 without applying anything, while the granular PUT /v1/config/{path} endpoint kept working. Add a dedicated runtime settings source ranked directly below the command line arguments and record granular updates there as well, so both endpoints share one store that survives re-evaluation of the settings sources. Environment variables keep precedence over the config file for all keys that were not set at runtime. Also repairs revert_settings() and update(), which passed their data through the same init settings. Closes #1303 * fix: env vars ignored on first config build ConfigEOS.__init__ passed self as first positional argument to _setup, which forwards it to pydantic_settings.BaseSettings.__init__. Its first positional parameter is _case_sensitive, so the environment source matched the upper case variable names against the lower case field names and returned nothing. Environment settings only took effect after the next configuration setup. * docs: changelog for config priority fixes * fix(config): preserve device identities and storage costs during migration * fix(measurement): restore JSON records into the existing singleton * fix(devices): preserve charge-rate typing and public import compatibility * feat(measurement): integrate typed energy quality and capacity APIs Port the locally backed-up measurement extensions to main async storage and PR #1256 device maps. Preserve runtime capacity estimates across #1305 bulk updates. Confirm JSON singleton restore defect on unchanged main and add regression. No production configuration or measurements included. Co-authored-by: Andreas <drbacke@gmx.de> * docs(measurement): describe household settings and consolidate regression coverage * docs(measurement): regenerate configuration and API contracts * test(measurement): isolate capacity database state between tests * ruff format fix * fix(measurement): restore JSON records into the existing singleton * test(measurement): assert restored timestamps before timezone conversion * test(measurement): assert restored timestamps before timezone conversion * fix: preserve imported feed-in revenue during parameter preparation Cancel GENETIC preparation when imported revenue cannot be read or contains invalid values, preserving the chosen provider instead of replacing it with demo tariffs. Keep valid positive, zero and negative amount/Wh series unchanged. Adapt the revenue-preservation regressions from PRs #1224 and #1304 to the async main API, including real timestamped imports and simulation repricing. The feature-only direct-marketing override remains outside this main fix. Co-authored-by: Christin <info@bikinibottom.capital> Co-authored-by: Normann <github@koldrack.com> * feat(devices): port slot-aware battery export and direct-use physics Port scoped device changes from |
||
|
|
3c862543a1 |
feat(devices): add bounded slot physics for GENETIC (#1327)
* feat: adapt configuration for multi optimization algorithms
Decouple configuration from optimization algorithm parameters. Add to_[algorithm]_param() methods
to the configuration that derive optimization algorithm specific parameters from the configuration.
Add x-scope tags to the configuration options that describe for which specific algorithms the
configuration option is for.
The whole device settings are restructured. There are now general settings for the device classes
with the afore mentioned to_[algorithm]_param() methods. The general device settings got their own
directory `devices/settings`. By this the parameter class also does not have to be a pydantic model
which can be used for future optimization/ simulations speed up.
Also the parameter class for a device is now part of the device module. This better decouples and
also is the natural place for parameters of a device.
Besides this feature there are also fixes and improvements:
* feat: extend home appliance time window settings and simulation
Home appliance can now be configured for multiple runs with per-cycle allowed time windows. The
number of remaining cycles to plan is determined at runtime by reading the
``cycles_completed_measurement_key`` from the measurement store.
* feat: specialiced CycleTimeWindowSequence for time window sequences
Sequence of time windows associated to cycles.
This model specializes ``ValueTimeWindowSequence`` so that the ``value``
field of each ``ValueTimeWindow`` encodes the **cycle index** (0-based
integer) the window belongs to.
Typical use: an appliance that must run ``n`` times per day, each run
constrained to a distinct time window. Assign ``value=0`` to windows
for the first cycle, ``value=1`` for the second, and so on. Multiple
windows may share the same cycle index (their allowed regions are unioned).
Windows with ``value=None`` are silently ignored by all cycle-aware methods.
* fix: Make test_configmigrate also regard the _ANY_SENTENIEL in key values
* chore: Make devices configurations a map instead of a list
This makes config paths stable regardless of declaration order and lets each device settings
class build its own config path from ``self.device_id`` without needing an external index.
Tests are adapted likewise.
Devices configurations are automatically migrated from lists to maps.
* chore: rename levelized_cost_of_storage_kwh to levelized_cost_of_storage_amt kwh
This better fits in the naming scheme and also makes clear the costs are money.
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com>
* fix: runtime config update ignored by config file
Runtime settings were handed back to pydantic-settings as init settings,
which rank below the config file and the environment. Any key already
present in EOS.config.json or in the environment silently discarded the
update, so a bulk PUT /v1/config returned 200 without applying anything,
while the granular PUT /v1/config/{path} endpoint kept working.
Add a dedicated runtime settings source ranked directly below the command
line arguments and record granular updates there as well, so both
endpoints share one store that survives re-evaluation of the settings
sources. Environment variables keep precedence over the config file for
all keys that were not set at runtime.
Also repairs revert_settings() and update(), which passed their data
through the same init settings.
Closes #1303
* fix: env vars ignored on first config build
ConfigEOS.__init__ passed self as first positional argument to _setup,
which forwards it to pydantic_settings.BaseSettings.__init__. Its first
positional parameter is _case_sensitive, so the environment source
matched the upper case variable names against the lower case field names
and returned nothing. Environment settings only took effect after the
next configuration setup.
* docs: changelog for config priority fixes
* fix(config): preserve device identities and storage costs during migration
* fix(devices): preserve charge-rate typing and public import compatibility
* ruff format fix
* feat(devices): port slot-aware battery export and direct-use physics
Port scoped device changes from
|
||
|
|
8224c64654 |
feat: integrate device and runtime configuration foundation on main (#1328)
* feat: adapt configuration for multi optimization algorithms Decouple configuration from optimization algorithm parameters. Add to_[algorithm]_param() methods to the configuration that derive optimization algorithm specific parameters from the configuration. Add x-scope tags to the configuration options that describe for which specific algorithms the configuration option is for. The whole device settings are restructured. There are now general settings for the device classes with the afore mentioned to_[algorithm]_param() methods. The general device settings got their own directory `devices/settings`. By this the parameter class also does not have to be a pydantic model which can be used for future optimization/ simulations speed up. Also the parameter class for a device is now part of the device module. This better decouples and also is the natural place for parameters of a device. Besides this feature there are also fixes and improvements: * feat: extend home appliance time window settings and simulation Home appliance can now be configured for multiple runs with per-cycle allowed time windows. The number of remaining cycles to plan is determined at runtime by reading the ``cycles_completed_measurement_key`` from the measurement store. * feat: specialiced CycleTimeWindowSequence for time window sequences Sequence of time windows associated to cycles. This model specializes ``ValueTimeWindowSequence`` so that the ``value`` field of each ``ValueTimeWindow`` encodes the **cycle index** (0-based integer) the window belongs to. Typical use: an appliance that must run ``n`` times per day, each run constrained to a distinct time window. Assign ``value=0`` to windows for the first cycle, ``value=1`` for the second, and so on. Multiple windows may share the same cycle index (their allowed regions are unioned). Windows with ``value=None`` are silently ignored by all cycle-aware methods. * fix: Make test_configmigrate also regard the _ANY_SENTENIEL in key values * chore: Make devices configurations a map instead of a list This makes config paths stable regardless of declaration order and lets each device settings class build its own config path from ``self.device_id`` without needing an external index. Tests are adapted likewise. Devices configurations are automatically migrated from lists to maps. * chore: rename levelized_cost_of_storage_kwh to levelized_cost_of_storage_amt kwh This better fits in the naming scheme and also makes clear the costs are money. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> * fix: runtime config update ignored by config file Runtime settings were handed back to pydantic-settings as init settings, which rank below the config file and the environment. Any key already present in EOS.config.json or in the environment silently discarded the update, so a bulk PUT /v1/config returned 200 without applying anything, while the granular PUT /v1/config/{path} endpoint kept working. Add a dedicated runtime settings source ranked directly below the command line arguments and record granular updates there as well, so both endpoints share one store that survives re-evaluation of the settings sources. Environment variables keep precedence over the config file for all keys that were not set at runtime. Also repairs revert_settings() and update(), which passed their data through the same init settings. Closes #1303 * fix: env vars ignored on first config build ConfigEOS.__init__ passed self as first positional argument to _setup, which forwards it to pydantic_settings.BaseSettings.__init__. Its first positional parameter is _case_sensitive, so the environment source matched the upper case variable names against the lower case field names and returned nothing. Environment settings only took effect after the next configuration setup. * docs: changelog for config priority fixes * fix(config): preserve device identities and storage costs during migration * fix(devices): preserve charge-rate typing and public import compatibility * ruff format fix * fix(config): satisfy typed device conversion and migration contracts * docs(config): refresh validated configuration prerequisite schemas --------- Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> Co-authored-by: Bobby Noelte <b0661n0e17e@gmail.com> Co-authored-by: r0b2g1t <r0b2g1t@users.noreply.github.com> |
||
|
AndreasandAndreas
|
cfa245b41a | fix: return only completed optimization results per run | ||
|
1abdd345c4 |
fix: unify mypy environments for local checks and CI (#1291)
The isolated pre-commit mypy hook previously omitted runtime type information that make mypy used, hiding errors involving dependencies such as Pydantic and Pendulum. Makefile, pre-commit and CI now run the same full-project typing policy in the development environment defined by uv.lock. - Use uv run --locked --exact --extra dev and the same mypy arguments for Makefile and the local hook. Check all of src and tests, including on configuration-only changes. - Pin Python 3.13 for local development and the pre-commit CI job, and install the locked pre-commit version in CI. - Disable incremental analysis because existing Pendulum cache state changes mypy 2.3.1 diagnostics. Document the policy, the performance tradeoff and the existing typing debt. - Add a regression test that exercises Makefile, the hook and the CI command in a temporary project, accepting valid dependency types and detecting deliberate Pydantic/Pendulum assignment errors. Resolve the newly detected mypy diagnostics. - Enable the numpydantic and Pydantic mypy plugins, retaining strict Pydantic constructor typing with init_typed = true. Validate raw/coercible payloads through model_validate. - Propagate concrete record, provider and time-window types through generic collections, factories and lookup methods. Preserve runtime field inspection and generated time-window documentation. - Align Pendulum annotations with actual factory/arithmetic results while retaining Pydantic validation adapters at runtime. Correct optional values, array boundaries, REST handlers and plotting interfaces. - Add pinned scipy-stubs and types-psutil, update uv.lock, and supply the plugins' dependencies. - Add runtime regression coverage for validated path defaults, normalized time-series metadata, generic field inspection, invalid timestamps and unsupported provider imports. Runtime and compatibility details: - Validate path defaults as Path objects while retaining raw string defaults needed by migration serialization with exclude_defaults. - Normalize feed-in tariff lists and default charge rates to NumPy arrays; reject missing timestamps/uninitialized values explicitly. Importing into a provider without import support returns HTTP 400. - Public JSON schemas and OpenAPI structure match main (excluding the generated version). Signed-off-by: dr-dimitry Signed-off-by: dr-dimitry Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> Co-authored-by: dr-dimitri <87113560+dr-dimitri@users.noreply.github.com> Co-authored-by: Normann <github@koldrack.com> |
||
|
Bobby NoelteandGitHub
|
ba76087db9 |
feat: electricity fee provider framework and generic providers (#1235)
Add new provider class for electricity fees providers. Add the generic providers: - ElecFeeFixed - ElecFeeImport The providers provide predictions for: - elecfee_consumption_amt_wh: Total fixed fee for consumed energy per Wh [amount/Wh]. This is the accumulation of all fixed per-Wh fees payable on "consumed energy - such as network charge, concession fee, and electricity charge - into a single amount. - elecfee_consumption_percent_amt: Total fixed surcharge on consumed energy, given as a percentage of the monetary amount already charged for that energy [%]. This is the accumulation of all percentage-based surcharges payable on top of the consumed-energy fee - such as VAT - into a single percentage. This is a percentage of the fee amount, not a per-Wh rate. - elecfee_feedin_amt_wh: Total fixed deduction from feed-in energy per Wh [amount/Wh]. This is the accumulation of all fixed per-Wh charges deducted from feed-in energy - such as metering fees or grid-operator handling "charges - into a single amount. Applied after the percentage-based deduction, i.e. it reduces the price by a flat amount per Wh rather than by a share of the raw price. - elecfee_feedin_percent_amt: Total percentage deducted from the raw feed-in price (spot price) [%]. This is the accumulation of all percentage-based deductions payable on the feed-in tariff - such as a marketing or balancing fee retained by the aggregator - into a single percentage. It is applied as `raw_price * (100 - percent) / 100`, i.e. it scales down the raw price rather than adding a surcharge to it. A new _apply_fee() method is added to the base class for ElecPrice and FeedInTariff to be used to add the fees in a consistent way. Fees are taken from the active ElecFee provider and applied to the raw prices given to the _apply_fee() method. The optional application of fees is added to: - ElecPriceAkkudoktor - ElecPriceFixed - ElecPriceEnergyCharts - ElecPriceSMARD - FeedInTariffEnergyCharts - FeedInTariffFixed - FeedInTariffSMARD The import providers ElecPriceImport and FeedInTariffImport do not apply fees by intentention. The following providers currently do not handle fees defined by ElecFee: - ElecPriceTibber - FeedInTariffAkkudoktor - FeedInTariffDvhubOnline - FeedInTariffTibber The tests for this feature are either added or existing tests are extended. The documentation was extended for the electricity fee provider settings. Besides this feature further improvements are added: * feat: add SMARD quarter-hour electricty price and feed-in tariff provider * feat: to_series method for TimeWindows and ValueTimeWindows Additional to to_array the time window sequence can now also produce a pandas series. Test have been extended to cover the series generation. * feat: use time windows in fixed feedin tariff provider Feedin tariff can now be configured by time windows - not a single value. * feat: EOSdash select for PVLib inverters and modules Provide PVLib inverter and module names in config selection. * feat: EOSdash lazy select for big option sets Add a new form for lazy selection of big option sets. Filtering and generation of the option set is done server-side. * fix: use raw data for ETS/ median prediction Use to raw time series data for ETS/ median prediction to avoid interference by e.g. dynamic grid charges. * fix: EOSdash config drops by type only on details resolve Drop configuration by type and path. Prevents dropping of configuration items with same type and level but different path. * fix: EOSdash configuration section closes on update Open section if searching or if last update touched this category — including updates on deeply nested sub-fields. * chore: make elecfeefixed, elecpricefixed and feedintarifffixed warn about no windows and default to 0 Missining configuration creates default 0 value and a warning instead of an exception. * fix: test setup for providers Reset db state on each test run. * chore: improve config option naming for elecpricefixed. * chore: adapt elecpricefixed test to changed time_windows naming * chore: factorized common price provider helpers to priceabc.py Factorized common price provider helpers to priceabc.py. Add tests for these helpers. Reduce/ change testing of elecpriceabc.py and feedintariffabc.py to cover only specifics. Rest of testing is already covered by test_priceabc.py. * chore: update version Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
886c93c92b |
fix: default server settings prevent env var config (#1234)
Change configuration source priorities to: - cli - environment vars - dotenv settings - config file settings - init settings By this the environment vars supersede any configuration var provided by the configuration file or by the initialisation with pydantic. The test_config.py::test_computed_path was fixed to to not use the defaul env var overwrite defined by conftest.py. This seemed to indicate non working env vars, but in fact was a test setupt fault. Besides this fix there are other fixes and changes added: * fix: exclude computed fields when merging settings Pydantic may overwrite settings by values given for computed fields and use these values instead of re-computing the field. Avoid computed fields in merging settings. * chore: improve Windows compatability of development setup Improve scripts to better run also on Windows. When doing path checks keep compatibility also to Windows pathes. A lot of changes to avoid the famous Windows CRLF handling and keep line endings to LF. * chore: add development hint for Windows Windows developers should set core.autocrlf to false. * chore: update version Signed-off-by: b0661 <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
269d2162a9 |
fix: server crash on prediction update failure (#1222)
Catch exceptions from prediction update and log as error message. Keep executing the energy management run. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
894790f577 |
feat: add pvlib pv forecast provider (#1214)
Add a PV forecast provider that calculates the forecast using a PVLib system model and weather forecast from the EOS weather forecast provider. Additional module and inverter models can be easily added as the database is build from PVLib and SAM databases and a bundled csv file. The module model and inververt model names are provided by new endpoints to be used in configuration. The provider is based on the fantastic work of EMHASS. See https://github.com/davidusb-geek/emhass/blob/master/src/emhass/forecast.py A short description of the provider is added to the documentation. Besides the new features there are the fixes and improvements: * feat: improve EOSdash config page * fix: kex_to_series for start_datetime Make key_to_series always start the series at start_datetime. * fix: default provider for GENETIC and GENETIC0 optimization To make the default less dependent on internet servers (with API changes and availability issues) the default for PVForecast is set to PVForecastPVLib and for ElecPrice to ElecPriceFixed. The default weather provider is changed to OpenMeteo. * fix: EOSdash display resampled prediction values Make EOSdash display resampled prediction values where resampling fits to the prediction value type. Use bar width that fits to 15 minutes value samples. * chore: add a UI hints system to EOSdash The UI hints system eases the definition of forms for configuration items. There are also forms for items in maps and lists. These forms allow to add and delete items to/ from maps and lists. The forms ensure that all required fields of newly added items are filled. * chore: Create an enum for valid optimization algorithms * chore. Make config also provide the available energy management modes. Used for configuration hints. * chore: Randomize default device id in configuration Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
 
|
9189fc890e |
fix(dataabc): drop NaN values in key_to_dict/key_to_lists (#1211)
The dropna filter compared values against float("nan") using ==, which is
always False (NaN != NaN). As a result NaN values were never dropped when
dropna=True, letting them leak into key_to_series/key_to_array and downstream
resampling.
Use pd.isna() to detect NaN, matching the rest of the module. Add regression
tests that fail before and pass after the fix.
Co-authored-by: Cornelius Mund <cornim@users.noreply.github.com>
Co-authored-by: Normann <github@koldrack.com>
|
||
|
|
b59012c1f7 |
fix(database): open backend so records actually persist and reload (#1209)
* fix(database): open backend so records actually persist and reload Measurement and prediction records were never persisted to the configured database backend (e.g. LMDB). They only survived while the process was running; every restart lost the accumulated history. For LoadAkkudoktorAdjusted this silently zeroed the measurement-based adjustment (the adjusted load forecast collapsed onto the raw mean). Root cause: `db_enabled` is defined as `database.is_open`, and every database code path (initialization, load, save, insert) is guarded behind `db_enabled`. Nothing ever opened the backend first -- the only lazy open (via `_run_db`) was unreachable because those calls sit behind the same guard. The backend therefore stayed closed, `db_enabled` stayed False, and all writes fell through to the JSON file fallback, which only holds the current in-memory snapshot and is not reloaded into the record store on startup. Fix: explicitly open the configured backend in `_db_ensure_initialized` before the `db_enabled` gate, wrapped in try/except so a failure degrades gracefully to file storage. Verified via an A/B test (identical persisted config, push -> save -> restart): without the fix the backend reports enabled=false and data is lost on restart; with the fix the backend is enabled, records persist to LMDB, and reload correctly after restart. * fix(database): skip disabled providers and avoid open retry loop Address review feedback: - Skip opening the backend for disabled providers (None/"NoDB"), whose is_open is always False and would otherwise be reopened on every call. - Add a one-shot _db_open_attempted flag so a closed/failed backend is not retried (and re-logged) on every record operation. - Add tests covering that NoDB never calls open() and that an unavailable backend does not raise per operation and is opened at most once. * fix(database): drop file-storage-fallback wording from open failure log --------- Co-authored-by: Cornelius Mund <cornim@users.noreply.github.com> |
||
|
Bobby NoelteandGitHub
|
1905682113 |
chore: adapt pdf visualization (#1205)
Change PDF visualization to be created on demand and per optimization algorithm. The PDF for the GENETIC0 optimization is provided by the /visualization_results.pdf endpoint. There is no change in the interface. By this the optimization algorithm is offloaded from the PDF generation which spares some time. To cope with several users may call the /visualization_results.pdf endpoint at the same time the PDF is generated on the fly without any intermediate file taking the stored GENETIC0 solution as an input. SVG picture generation is removed as this would again create intermediate files. Chart pictures can easily be taken from the PDF. To allow on demand creation of the optimization results visualization the optimisation solution stored is extended by several new attributes. To keep the deprecated /optimize endpoint compatible the optimization solution is stripped to the legacy content before returned. Due to the extension of the solution the optimization tests were adapted to cover the extended content. The optimization tests are adapted to test the generated visualization report by the pypdf reader. Pypdf is added to the development dependencies. Besides the adaptation several fixes and improvements are added: * feat: extend /v1/prediction/series endpoint by resampling and filling Add parameters for resampling and filling. Add the processing parameter to control wether raw data or resampled data shall be returned. * feat: extend /v1/measurement/series endpoint by resampling and filling Add parameters for resampling and filling: Add the processing parameter to control wether raw data or resampled data shall be returned. * feat: standardize and improve API error response Use FASTApi exception handlers to provide a standardized API exception handling. All exceptions are logged. Exception traces are only returned if the new logging configuration parameter logging.api_logging_level is set to "DEBUG" or "TRACE". Avoids unwanted leackage of server internals on exceptions. * fix: align to intervall when resampling Ensure resampling is aligned to interval also when the buckets are shifted due to the align_to_intervall parameter is set. * chore: make dropna mandatory and default to True * chore: refactor key_to_xxx data management methods Make key_to_series the central method for data resampling and fill. Add a new key_to_raw_series to retrieve the data as it is stored (without resampling and filling). Users of key_to_series were mostly moved to key_to_raw_series as this resembles the former interface. Especially in predictions and tests this was done. * chore: create test data sub-directory for each optimization algorithm To prevent cluttering the test data directory and ease test data management for optimization algorithms each algorithm got it's own sub-directory. The current test data was moved to these sub-directories. * chore: update version Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
8344974c16 |
chore: adapt deprecated endpoints to 15-minutes predictions (#1195)
Bump Version / Bump Version Workflow (push) Canceled after 0s
CodeQL Advanced / Analyze (actions) (push) Canceled after 0s
CodeQL Advanced / Analyze (python) (push) Canceled after 0s
docker-build / platform-excludes (push) Canceled after 0s
pre-commit / pre-commit (push) Canceled after 0s
Run Pytest on Pull Request / test (push) Canceled after 0s
docker-build / build (push) Canceled after 0s
docker-build / merge (push) Canceled after 0s
Ensure the deprecated endpoints get /strompreis, post /gesamtlast, get /gesamtlast_simple,
get /pvforecast to work on 1-hour intervalls even if the prediction provides 15-minutes
intervall data. This keeps the interface compliant to the legacy functionality.
Besides this adaptations several other improvements and fixes are included in this PR.
* feat: extend data management method key_to array by resample_method
One can now define the resample method on how to aggregate the values in an interval
for resampling. Three methods are provided:
- "first": Use the first value in each interval.
- "mean": Compute the arithmetic mean of all samples in each interval.
- "interval_mean": Compute the time-weighted mean assuming each value remains valid
until the next timestamp (piecewise-constant signal).
* feat: extend the get /v1/prediction/dataframe endpoint with resampling parameters
Make all parameters for resampling available at the endpoint.
* feat: extend the get /v1/prediction/list endpoint with resampling parameters
Make all parameters for resampling available at the endpoint.
* feat: add new delete /v1/prediction/range endpoint
The endpoint allows to delete prediction values for a given time span.
* fix: adapt for PVForecastAkkudoktor server side cache handling
/api.akkudoktor/forecast does it's own caching on requests. Call it with slightly
randomized request values to avoid getting cached values in the case we need fresh
data. The requests are anyway rate limited to one request per hour on our side.
* chore: add core.types
This module centralizes reusable type definitions shared across multiple
packages. Defining common types here avoids duplication of complex type
annotations (such as Literal aliases), ensures consistent typing across the
code base, and helps prevent circular import dependencies between modules.
* chore: extend cache testing
* chore: add system test for deprecated /strompreis endpoint
* chore: add unit test module for server endpoints
Add a new test module to do unit tests on server endpoints. First test added
for deprecated get /strompreis endpoint.
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com>
|
||
|
Bobby NoelteandGitHub
|
e23bb7b497 |
chore: prepare for update of genetic algorithm (#1190)
Bump Version / Bump Version Workflow (push) Has been cancelled
CodeQL Advanced / Analyze (actions) (push) Has been cancelled
CodeQL Advanced / Analyze (python) (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Andreas will update the genetic algorithm for 15-minutes optimization intervals. Copy the current GENETIC optimization algorithm to GENETIC0 to enable to keep the algorithm with the current functionality. Also copy resources like the load interpolator to the GENETIC0 algorithm to keep them despite possible later changes to the interpolator. Make the deprecated legacy /optimize endpoint use the GENETIC0 optimization algorithm to in-fact behave the same way even if there will later be changes to the GENETIC algorithm by Andreas. Add a new REST endpoint to provide the unprocessed optimisation results of the GENETIC and GENETIC0 algorithm in case one wants to use them as done with the deprecated /optimize endpoint. Adapt the optimization configuration to have distinct configurations for the GENETIC and the GENETIC0 algorithm. Create a copy of the current tests for the GENETIC algorithm to be used for the GENETIC0 algorithm. This avoids the tests for the GENETIC0 algorithm to be influenced by later changes by Andreas. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
2ed04d5573 |
fix: db compaction run (#1156)
These fixes were taken from https://github.com/arneman/EOS/tree/refactor/economic-objective. 1. Fix compaction job registration bug in eos.py: - compact_eos_database was registered with save_eos_database function - Now correctly calls compact_eos_database() - Root cause: compaction/vacuum never ran, allowing records to grow unbounded 2. Optimize db_iterate_records() from O(n) to O(log n): - Previous: linear scan from index 0 for every key_to_array() call - Now: uses bisect_left on _db_sorted_timestamps to skip to start position - Critical for large datasets: 9400 measurement records → 94s/call overhead → 5 calls/optimization 3. Reduce compaction_interval_sec default from 604800s (7 days) to 3600s (1 hour): - With 7-day interval: 65000 records accumulate before first cleanup - Bridge pushes ~1 record/9s (grid_export_emr) - Hourly compaction + 2h data window → stable ~950 records Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
 |
8bf798daeb |
fix: bare except (#1159)
Bare `except:` catches SystemExit and KeyboardInterrupt, which can: - Mask Ctrl+C handling during long optimization runs - Hide critical system signals - Make debugging harder Replacement by `except Exception:` preserves the same error handling while respecting system-level exceptions. Co-authored-by: Milo @KeloYuan Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
 
|
6933b33542 |
feat: rename genetic optimization API fields to English (#675)
Bump Version / Bump Version Workflow (push) Has been cancelled
CodeQL Advanced / Analyze (actions) (push) Has been cancelled
CodeQL Advanced / Analyze (python) (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Rename the German field names of the genetic optimization API to English with full backward compatibility: - English names are canonical and documented in the OpenAPI schema; the German names are still accepted on input via validation aliases and re-emitted in responses as deprecated computed fields - fix visualization receiving the raw German-keyed simulation dict (KeyError: load_wh_per_hour) - rename internal German identifiers (optimize_ems, total_balance, battery_residual_value, self_consumption, extra_data keys) - use amount instead of currency/euro in chart labels and schema descriptions; document currency handling (whole currency units, never cents) in the API description - regenerate openapi.json and generated docs Co-authored-by: Tobias Welz <tobias.wizneteu@gmail.com> |
||
|
Bobby NoelteandGitHub
|
eb9e966de9 |
fix: move data management to async (#1015)
FAstAPI is an async framework. Data may be imported and exported, load and save, set and get asynchronously. Prevent interleaving data operations to corrupt the data. In the previous design sync and async data access was intermixed leading to data corruption. The basic data classes DataSequence and DataContainer and the derived classes like Provider and Measurement now are async. Data access is protected by several async locks. To support the async design of the data classes the database interface became async. The energy management is also adapted to the new async design. Optimization is still off-loaded to another thread, but the prepration for the optimization and the post optimization actions now follow the async design. Adapter operations are now also protected by async locks. Tests were adapted to the async design and new tests were created. Besides this major fix several other improvements and fixes are included in this PR. * fix: key_to_dict/list/array only regard data records with key value set. Before the exclusion of no value data records was only done if the dropna flag was set. * fix: test for visual result pdf generation Due to updates in the library the generated charts text was a little bit different. Adapt the test to create the comaprison pdf in the test data durectory and update the reference pdf. * chore: Remove MutableMapping from DataSequence and DataContainer. Mutable Mapping does not fit to the now async design. * chore: Add NoDB database backend This backend implements the full database backend interface but performs no actual persistence. It is intended for configurations where database persistence is disabled (`provider=None`). * chore: Improve measurement data import testing with real world scenarios. Added two new endpoints to support testing. * chore: Add mermaid to supported documentation tools * chore: Add documentation about async design * chore: Add documentation about generic data handling Covers the basics of measurement and prediction time series data handling. * chore: Add empty lines around markdown lists. * chore: sync pre-commit config to updated package versions Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
8c7f7f321d |
fix: catch AttributeError in validate_data to allow Union fallthrough for non-DataFrame inputs (#993)
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Bump Version / Bump Version Workflow (push) Has been cancelled
CodeQL Advanced / Analyze (actions) (push) Has been cancelled
CodeQL Advanced / Analyze (python) (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
* fix: catch AttributeError in validate_data for non-DataFrame Union types
PUT /v1/prediction/import/{provider_id} crashed with AttributeError
when Union validator tried PydanticDateTimeDataFrame validation on
scalar float/int values. Instead of falling through to DateTimeData
or dict, the exception propagated as 500. Fixes #992.
* fix: correct indentation of ValueError in validate_data
---------
Co-authored-by: Normann <github@koldrack.com>
|
||
|
Bobby NoelteandGitHub
|
5fb5234e7d |
fix: device id in solution follows configuration (#964)
Bump Version / Bump Version Workflow (push) Has been cancelled
CodeQL Advanced / Analyze (actions) (push) Has been cancelled
CodeQL Advanced / Analyze (python) (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Make device id in solution follow actual configuration. Adapt version update to new CI bump workflow design. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
9f74c4a515 |
chore: Prepare Release v0.3.0 (#959)
This pull request prepares release **v0.3.0**. Akkudoktor-EOS can now be run as Home Assistant add-on and standalone. As Home Assistant add-on EOS uses ingress to fully integrate the EOSdash dashboard in Home Assistant. Adapters for Home Assistant and NodeRed integration are added. These adapters provide a simplified interface to these HEMS besides the standard REST interface. The prediction and measurement data can now be backed by a database. The database allows to keep historic prediction data and measurement data for long time without keeping it in memory. The database supports backend selection, compression, incremental data load, automatic data saving to storage, automatic vacuum and compaction. Two database backends are integrated and can be configured, LMDB and SQLight3. New prediction providers allow to access OpenMeteo weather data and to define fixed electricity prices for configurable time windows. An anoying bug in the genetic algorithm that created unfeasable battery charge and discharge amounts is now hopefully fixed. In addition, bugs were fixed and new features were added. See CHANGELOG for full details. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
71e5abce88 |
fix: energy charts bidding zone in request (#948)
Bump Version / Bump Version Workflow (push) Has been cancelled
CodeQL Advanced / Analyze (actions) (push) Has been cancelled
CodeQL Advanced / Analyze (python) (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Ensure that the bidding zone in the request is correctly set to a string value (not an enum). This seems to be also an issue with python version < 3.11. Add safeguards to only use python >= 3.11. Still keep a regression test for the enum conversion to string. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
b67148c47f |
feat: make home assistant add-on run optimization by default (#940)
When running as Home Assistant add-on the only viable usage is running with cyclic optimization. Make this the default to als propvide a better experience for first time users. The optimization will start with demo data, which also helps to configure Akkudoktor-EOS to the personal usage. The disabling of the automatic energy management is now an explicit mode "DISABLED". Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
cf477d91a3 |
feat: add fixed electricity prediction with time window support (#930)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Add a fixed electricity prediction that supports prices per time window.
The time windows may flexible be defined by day or date.
The prediction documentation is updated to also cover the ElecPriceFixed
provider.
The feature includes several changes that are not directly related to the
electricity price prediction implementation but are necessary to keep
EOS running properly and to test and document the changes.
* feat: add value time windows
Add time windows with an associated float value.
* feat: harden eos measurements endpoints error detection and reporting
Cover more errors that may be raised during endpoint access. Report the
errors including trace information to ease debugging.
* feat: extend server configuration to cover all arguments
Make the argument controlled options also available in server configuration.
* fix: eos config configuration by cli arguments
Move the command line argument handling to config eos so that it is
excuted whenever eos config is rebuild or reset.
* chore: extend measurement endpoint system test
* chore: refactor time windows
Move time windows to configabc as they are only used in configurations.
Also move all tests to test_configabc.
* chore: provide config update errors in eosdash with summarized error text
If there is an update error provide the error text as a summary. On click
provide the full error text.
* chore: force eosdash ip address and port in makefile dev run
Ensure eosdash ip address and port are correctly set for development runs.
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com>
|
||
|
Bobby NoelteandGitHub
|
1832364d60 |
fix: pydantic datetime dataframe timezone conversion (#927)
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
997e7646e9 |
fix: prevent exception when load prediction data is missing (#925)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Validate solution prediction data before processing. If required prediction data is missing, the prediction is skipped instead of raising an exception. Introduce a new configuration file saving policy to improve loading robustness: - Exclude computed fields - Exclude fields set to their default values - Exclude fields with value None - Use field aliases - Recursively remove empty dictionaries and lists - Ensure general.version is always present and correctly set When loading older configuration files, computed fields are now stripped before migration. This further improves backward compatibility and loading robustness. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
04420e66ab |
fix: Improve provider update error handling and add VRM provider settings validation (#887)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
* fix: improve error handling for provider updates Distinguishes failures of active providers from inactive ones. Propagates errors only for enabled providers, allowing execution to continue if a non-active provider fails, which avoids unnecessary interruptions and improves robustness. * fix: add provider settings validation for forecast requests Prevents potential runtime errors by checking if provider settings are configured before accessing forecast credentials. Raises a clear error when settings are missing to help with debugging misconfigurations. * refactor(load): move provider settings to top-level fields Transitions load provider settings from a nested "provider_settings" object with provider-specific keys to dedicated top-level fields.\n\nRemoves the legacy "provider_settings" mapping and updates migration logic to ensure backward compatibility with existing configurations. * docs: update version numbers and documantation --------- Co-authored-by: Normann <github@koldrack.com> |
||
|
Bobby NoelteandGitHub
|
32e690becf |
fix: EOS run asynchronous tasks (#904)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Startup retention manager for asynchronous tasks. Handle gracefully exceptions in these tasks or the configuration for them. Remove tasks.py as repeated tasks are now handled by the retention manager. When running on GitHub, only the version date file is checked. The development tag is merely a label, so any date set during development suffices. The test_doc is also skipped on GitHub actions. |
||
|
Bobby NoelteandGitHub
|
378377b1ce |
fix: test break docs and on data compaction (2) (#902)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Ensure that the snapping sequence generated in the test fixture is within the boundaries expected by the test. Ensure we read the _version_date.py info as UTC datetime and do no localtime conversion. Prevent and guard test_version.py to modify the version date file. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
90e2e8af7e |
fix: test break on docs version and data compaction (#900)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Upgrade to: - pandas==3.01 - fastapi[standard-no-fastapi-cloud-cli]==0.132.0 - fastapi_cli==0.0.23 - MonsterUI==1.0.44 - uvicorn==0.41.0 Close database in database fixture on teardown. Fix file exclusion in hash and version date generation. Update version information in documentation. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
d446274129 |
fix: Adapt versioning scheme to Home Assistant and switch to uv (#896)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
Home Assistant expects versioning always increases numbers. Add a date component to the development version to comply with this expectation. The scheme is now 0.0.0.dev<date><hash>. Use uv for creating and managing the virtual environment for developement. This enourmously speeds up dependency updates. For this change dependency requirements are now solely handled in pyproject.toml. requirements.tx and requirements-dev.txt are deleted. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
6498c7dc32 |
Add database support for measurements and historic prediction data. (#848)
The database supports backend selection, compression, incremental data load, automatic data saving to storage, automatic vaccum and compaction. Make SQLite3 and LMDB database backends available. Update tests for new interface conventions regarding data sequences, data containers, data providers. This includes the measurements provider and the prediction providers. Add database documentation. The fix includes several bug fixes that are not directly related to the database implementation but are necessary to keep EOS running properly and to test and document the changes. * fix: config eos test setup Make the config_eos fixture generate a new instance of the config_eos singleton. Use correct env names to setup data folder path. * fix: startup with no config Make cache and measurements complain about missing data path configuration but do not bail out. * fix: soc data preparation and usage for genetic optimization. Search for soc measurments 48 hours around the optimization start time. Only clamp soc to maximum in battery device simulation. * fix: dashboard bailout on zero value solution display Do not use zero values to calculate the chart values adjustment for display. * fix: openapi generation script Make the script also replace data_folder_path and data_output_path to hide real (test) environment pathes. * feat: add make repeated task function make_repeated_task allows to wrap a function to be repeated cyclically. * chore: removed index based data sequence access Index based data sequence access does not make sense as the sequence can be backed by the database. The sequence is now purely time series data. * chore: refactor eos startup to avoid module import startup Avoid module import initialisation expecially of the EOS configuration. Config mutation, singleton initialization, logging setup, argparse parsing, background task definitions depending on config and environment-dependent behavior is now done at function startup. * chore: introduce retention manager A single long-running background task that owns the scheduling of all periodic server-maintenance jobs (cache cleanup, DB autosave, …) * chore: canonicalize timezone name for UTC Timezone names that are semantically identical to UTC are canonicalized to UTC. * chore: extend config file migration for default value handling Extend the config file migration handling values None or nonexisting values that will invoke a default value generation in the new config file. Also adapt test to handle this situation. * chore: extend datetime util test cases * chore: make version test check for untracked files Check for files that are not tracked by git. Version calculation will be wrong if these files will not be commited. * chore: bump pandas to 3.0.0 Pandas 3.0 now performs inference on the appropriate resolution (a.k.a. unit) for the output dtype which may become datetime64[us] (before it was ns). Also numeric dtype detection is now more strict which needs a different detection for numerics. * chore: bump pydantic-settings to 2.12.0 pydantic-settings 2.12.0 under pytest creates a different behaviour. The tests were adapted and a workaround was introduced. Also ConfigEOS was adapted to allow for fine grain initialization control to be able to switch off certain settings such as file settings during test. * chore: remove sci learn kit from dependencies The sci learn kit is not strictly necessary as long as we have scipy. * chore: add documentation mode guarding for sphinx autosummary Sphinx autosummary excecutes functions. Prevent exceptions in case of pure doc mode. * chore: adapt docker-build CI workflow to stricter GitHub handling Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
58d70e417b |
feat: add Home Assistant and NodeRED adapters (#764)
Adapters for Home Assistant and NodeRED integration are added. Akkudoktor-EOS can now be run as Home Assistant add-on and standalone. As Home Assistant add-on EOS uses ingress to fully integrate the EOSdash dashboard in Home Assistant. The fix includes several bug fixes that are not directly related to the adapter implementation but are necessary to keep EOS running properly and to test and document the changes. * fix: development version scheme The development versioning scheme is adaptet to fit to docker and home assistant expectations. The new scheme is x.y.z and x.y.z.dev<hash>. Hash is only digits as expected by home assistant. Development version is appended by .dev as expected by docker. * fix: use mean value in interval on resampling for array When downsampling data use the mean value of all values within the new sampling interval. * fix: default battery ev soc and appliance wh Make the genetic simulation return default values for the battery SoC, electric vehicle SoC and appliance load if these assets are not used. * fix: import json string Strip outer quotes from JSON strings on import to be compliant to json.loads() expectation. * fix: default interval definition for import data Default interval must be defined in lowercase human definition to be accepted by pendulum. * fix: clearoutside schema change * feat: add adapters for integrations Adapters for Home Assistant and NodeRED integration are added. Akkudoktor-EOS can now be run as Home Assistant add-on and standalone. As Home Assistant add-on EOS uses ingress to fully integrate the EOSdash dashboard in Home Assistant. * feat: allow eos to be started with root permissions and drop priviledges Home assistant starts all add-ons with root permissions. Eos now drops root permissions if an applicable user is defined by paramter --run_as_user. The docker image defines the user eos to be used. * feat: make eos supervise and monitor EOSdash Eos now not only starts EOSdash but also monitors EOSdash during runtime and restarts EOSdash on fault. EOSdash logging is captured by EOS and forwarded to the EOS log to provide better visibility. * feat: add duration to string conversion Make to_duration to also return the duration as string on request. * chore: Use info logging to report missing optimization parameters In parameter preparation for automatic optimization an error was logged for missing paramters. Log is now down using the info level. * chore: make EOSdash use the EOS data directory for file import/ export EOSdash use the EOS data directory for file import/ export by default. This allows to use the configuration import/ export function also within docker images. * chore: improve EOSdash config tab display Improve display of JSON code and add more forms for config value update. * chore: make docker image file system layout similar to home assistant Only use /data directory for persistent data. This is handled as a docker volume. The /data volume is mapped to ~/.local/share/net.akkudoktor.eos if using docker compose. * chore: add home assistant add-on development environment Add VSCode devcontainer and task definition for home assistant add-on development. * chore: improve documentation |
||
|
Bobby NoelteandGitHub
|
976a2c8405 |
chore: automate development version and release generation (#772)
Bump Version / Bump Version Workflow (push) Has been cancelled
docker-build / platform-excludes (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
This change introduces a GitHub Action to automate release creation, including proper tagging and automatic addition of a development marker to the version. A hash is also appended to development versions to make their state easier to distinguish. Tests and release documentation have been updated to reflect the revised release workflow. Several files now retrieve the current version dynamically. The test --full-run option has been rename to --finalize to make clear it is to do commit finalization testing. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
7bf9dd723e |
chore: improve doc generation and test (#762)
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
Improve documentation generation and add tests for documentation. Extend sphinx by todo directive. The configuration table is now split into several tables. The test is adapted accordingly. There is a new test that checks the docstrings to be compliant to the RST format as used by sphinx to create the documentation. We can not use Markdown in docstrings. The docstrings are adapted accordingly. An additional test checks that the documentation can be build with sphinx. This test takes very long is only enabled in full run (aka. ci) mode. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby Noelte
|
8da137f8f1 |
fix: cached_method deprecated and test
cachebox deprecated the method decorator. Used cached instead. Fix cache integration tests that were accessing real world addresses. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby NoelteandGitHub
|
e7b43782a4 |
fix: pydantic extra keywords deprecated (#753)
Pydantic deprecates using extra keyword arguments on Field. Used json_schema_extra instead. Deprecated in Pydantic V2.0 to be removed in V3.0. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |
||
|
Bobby Noelte
|
54b0622a96 |
chore: set development version marker 0.2.0+dev
docker-build / platform-excludes (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled
Run Pytest on Pull Request / test (push) Has been cancelled
docker-build / build (push) Has been cancelled
docker-build / merge (push) Has been cancelled
Close stale pull requests/issues / Find Stale issues and PRs (push) Has been cancelled
* Development version v0.2.0+dev This pull request marks the repository as back in active development. * Changes - Set version to `v0.2.0+dev` No changelog entry is needed. Signed-off-by: Bobby Noelte <b0661n0e17e@gmail.com> |