feat(measurement): add typed energy, quality and capacity APIs (#1326)

* 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

* test(measurement): assert restored timestamps before timezone conversion

* docs(measurement): align API version with refreshed prerequisites

* fix(config): satisfy typed device conversion and migration contracts

* docs(config): refresh validated configuration prerequisite schemas

* fix(measurement): enforce typed capacity and sample validation

* style(measurement): normalize imports for CI

* docs(measurement): refresh typed measurement API 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>
This commit is contained in:
Andreas
2026-09-17 18:28:43 +02:00
committed by GitHub
co-authored by Andreas Bobby Noelte r0b2g1t
parent 8224c64654
commit 1a18935667
17 changed files with 2972 additions and 5 deletions
+8
View File
@@ -35,6 +35,8 @@ config path from ``self.device_id`` without needing an external index.
"batteries": {
"bat0": {
"device_id": "bat0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 8000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
@@ -62,6 +64,8 @@ config path from ``self.device_id`` without needing an external index.
"electric_vehicles": {
"ev0": {
"device_id": "ev0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 60000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
@@ -116,6 +120,8 @@ config path from ``self.device_id`` without needing an external index.
"batteries": {
"bat0": {
"device_id": "bat0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 8000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
@@ -155,6 +161,8 @@ config path from ``self.device_id`` without needing an external index.
"electric_vehicles": {
"ev0": {
"device_id": "ev0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 60000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
+7
View File
@@ -39,6 +39,8 @@
"batteries": {
"bat0": {
"device_id": "bat0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 8000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
@@ -66,6 +68,8 @@
"electric_vehicles": {
"ev0": {
"device_id": "ev0",
"capacity_estimation": null,
"capacity_estimate": null,
"capacity_wh": 60000,
"charging_efficiency": 0.88,
"discharging_efficiency": 0.88,
@@ -205,6 +209,9 @@
},
"measurement": {
"historic_hours": 17520,
"channels": {},
"household": null,
"energy_context_seconds": 86400,
"load_emr_keys": [
"load0_emr"
],
+9
View File
@@ -7,9 +7,12 @@
| Name | Environment Variable | Type | Read-Only | Default | Description |
| ---- | -------------------- | ---- | --------- | ------- | ----------- |
| channels | `EOS_MEASUREMENT__CHANNELS` | `dict[str, akkudoktoreos.measurement.measurement.MeasurementChannelSettings]` | `rw` | `required` | Typed raw measurement channels keyed by measurement key. |
| energy_context_seconds | `EOS_MEASUREMENT__ENERGY_CONTEXT_SECONDS` | `int` | `rw` | `86400` | None |
| grid_export_emr_keys | `EOS_MEASUREMENT__GRID_EXPORT_EMR_KEYS` | `Optional[list[str]]` | `rw` | `None` | The keys of the measurements that are energy meter readings of energy export to grid [kWh]. |
| grid_import_emr_keys | `EOS_MEASUREMENT__GRID_IMPORT_EMR_KEYS` | `Optional[list[str]]` | `rw` | `None` | The keys of the measurements that are energy meter readings of energy import from grid [kWh]. |
| historic_hours | `EOS_MEASUREMENT__HISTORIC_HOURS` | `Optional[int]` | `rw` | `17520` | Number of hours into the past for measurement data |
| household | `EOS_MEASUREMENT__HOUSEHOLD` | `Optional[akkudoktoreos.measurement.household.HouseholdSettings]` | `rw` | `None` | Optional household energy balance definition. |
| keys | | `list[str]` | `ro` | `N/A` | The keys of the measurements that can be stored. |
| load_emr_keys | `EOS_MEASUREMENT__LOAD_EMR_KEYS` | `Optional[list[str]]` | `rw` | `None` | The keys of the measurements that are energy meter readings of a load [kWh]. |
| pv_production_emr_keys | `EOS_MEASUREMENT__PV_PRODUCTION_EMR_KEYS` | `Optional[list[str]]` | `rw` | `None` | The keys of the measurements that are PV production energy meter readings [kWh]. |
@@ -25,6 +28,9 @@
{
"measurement": {
"historic_hours": 17520,
"channels": {},
"household": null,
"energy_context_seconds": 86400,
"load_emr_keys": [
"load0_emr"
],
@@ -51,6 +57,9 @@
{
"measurement": {
"historic_hours": 17520,
"channels": {},
"household": null,
"energy_context_seconds": 86400,
"load_emr_keys": [
"load0_emr"
],
+172 -1
View File
@@ -1,6 +1,6 @@
# Akkudoktor-EOS
**Version**: `v0.3.0.dev2609171139845429`
**Version**: `v0.3.0.dev2609171537448705`
<!-- pyml disable line-length -->
**Description**: This project provides a comprehensive solution for simulating and optimizing an energy system based on renewable energy sources. With a focus on photovoltaic (PV) systems, battery storage (batteries), load management (consumer requirements), heat pumps, electric vehicles, and consideration of electricity price data, this system enables forecasting and optimization of energy flow and costs over a specified period.
@@ -880,6 +880,44 @@ Returns:
---
## POST /v1/measurement/battery-capacity/{battery_id}
<!-- pyml disable line-length -->
**Links**: [local](http://localhost:8503/docs#/default/post_battery_capacity_v1_measurement_battery-capacity__battery_id__post), [eos](https://petstore3.swagger.io/?url=https://raw.githubusercontent.com/Akkudoktor-EOS/EOS/refs/heads/main/openapi.json#/default/post_battery_capacity_v1_measurement_battery-capacity__battery_id__post)
<!-- pyml enable line-length -->
Post Battery Capacity
<!-- pyml disable line-length -->
```python
"""
Estimate capacity from independent SoC anchors and configured DC power.
store_estimate writes the separate capacity_estimate config field in memory.
Persistence follows the regular EOS configuration save mechanism.
The active capacity_wh and raw measurements are never changed here.
"""
```
<!-- pyml enable line-length -->
**Parameters**:
- `battery_id` (path, required): No description provided.
**Request Body**:
- `application/json`: {
"$ref": "#/components/schemas/BatteryCapacityRequest"
}
**Responses**:
- **200**: Successful Response
- **422**: Validation Error
---
## PUT /v1/measurement/data
<!-- pyml disable line-length -->
@@ -940,6 +978,72 @@ Merge the measurement data given as dataframe into EOS measurements.
---
## GET /v1/measurement/energy
<!-- pyml disable line-length -->
**Links**: [local](http://localhost:8503/docs#/default/get_energy_v1_measurement_energy_get), [eos](https://petstore3.swagger.io/?url=https://raw.githubusercontent.com/Akkudoktor-EOS/EOS/refs/heads/main/openapi.json#/default/get_energy_v1_measurement_energy_get)
<!-- pyml enable line-length -->
Get Energy
<!-- pyml disable line-length -->
```python
"""
Energy in Wh, with temporal coverage and quality; no missing-to-zero filling.
"""
```
<!-- pyml enable line-length -->
**Parameters**:
- `key` (query, required): No description provided.
- `start` (query, required): No description provided.
- `end` (query, required): No description provided.
- `interval_seconds` (query, optional): No description provided.
**Responses**:
- **200**: Successful Response
- **422**: Validation Error
---
## GET /v1/measurement/household
<!-- pyml disable line-length -->
**Links**: [local](http://localhost:8503/docs#/default/get_household_v1_measurement_household_get), [eos](https://petstore3.swagger.io/?url=https://raw.githubusercontent.com/Akkudoktor-EOS/EOS/refs/heads/main/openapi.json#/default/get_household_v1_measurement_household_get)
<!-- pyml enable line-length -->
Get Household
<!-- pyml disable line-length -->
```python
"""
Site, household without EV, and base without separately measured devices.
"""
```
<!-- pyml enable line-length -->
**Parameters**:
- `start` (query, required): No description provided.
- `end` (query, required): No description provided.
- `interval_seconds` (query, optional): No description provided.
**Responses**:
- **200**: Successful Response
- **422**: Validation Error
---
## GET /v1/measurement/keys
<!-- pyml disable line-length -->
@@ -994,6 +1098,73 @@ Delete measurement values for a key within a datetime range.
---
## GET /v1/measurement/samples
<!-- pyml disable line-length -->
**Links**: [local](http://localhost:8503/docs#/default/get_samples_v1_measurement_samples_get), [eos](https://petstore3.swagger.io/?url=https://raw.githubusercontent.com/Akkudoktor-EOS/EOS/refs/heads/main/openapi.json#/default/get_samples_v1_measurement_samples_get)
<!-- pyml enable line-length -->
Get Samples
<!-- pyml disable line-length -->
```python
"""
Read raw samples including quality, in a bounded half-open range.
"""
```
<!-- pyml enable line-length -->
**Parameters**:
- `key` (query, required): No description provided.
- `start` (query, required): No description provided.
- `end` (query, required): No description provided.
**Responses**:
- **200**: Successful Response
- **422**: Validation Error
---
## PUT /v1/measurement/samples
<!-- pyml disable line-length -->
**Links**: [local](http://localhost:8503/docs#/default/put_samples_v1_measurement_samples_put), [eos](https://petstore3.swagger.io/?url=https://raw.githubusercontent.com/Akkudoktor-EOS/EOS/refs/heads/main/openapi.json#/default/put_samples_v1_measurement_samples_put)
<!-- pyml enable line-length -->
Put Samples
<!-- pyml disable line-length -->
```python
"""
Upsert raw values and their quality; legacy value/series payloads remain valid.
"""
```
<!-- pyml enable line-length -->
**Request Body**:
- `application/json`: {
"type": "array",
"items": {
"$ref": "#/components/schemas/MeasurementSample"
},
"maxItems": 10000,
"title": "Samples"
}
**Responses**:
- **200**: Successful Response
- **422**: Validation Error
---
## GET /v1/measurement/series
<!-- pyml disable line-length -->
+965 -1
View File
File diff suppressed because it is too large Load Diff
@@ -11,6 +11,10 @@ from numpydantic import NDArray, Shape
from pydantic import Field, computed_field, field_validator, model_validator
from akkudoktoreos.devices.settings.devicebasesettings import DevicesBaseSettings
from akkudoktoreos.measurement.batterycapacity import (
BatteryCapacityEstimate,
BatteryCapacityEstimationSettings,
)
if TYPE_CHECKING:
from akkudoktoreos.devices.genetic0.genetic0battery import (
@@ -35,6 +39,12 @@ class BatteriesCommonSettings(DevicesBaseSettings):
Note: Used for the GENETIC and GENETIC0 algorithm.
"""
capacity_estimation: Optional[BatteryCapacityEstimationSettings] = None
capacity_estimate: Optional[BatteryCapacityEstimate] = Field(
default=None,
description="Last explicitly stored capacity estimate; does not override capacity_wh.",
)
capacity_wh: int = Field(
default=8000,
gt=0,
@@ -0,0 +1,213 @@
"""Estimate effective model capacity from independent SoC anchors and DC power."""
from datetime import datetime
from math import isfinite
from typing import TYPE_CHECKING, Literal
from pydantic import AwareDatetime, BaseModel, ConfigDict, Field, model_validator
from akkudoktoreos.config.configabc import SettingsBaseModel
from akkudoktoreos.measurement.quality import SampleQuality
if TYPE_CHECKING:
from akkudoktoreos.measurement.measurement import MeasurementChannelSettings
class BatteryCapacityEstimationSettings(SettingsBaseModel):
"""Only a battery-terminal DC power channel is supported, not inverter AC."""
power_key: str = Field(min_length=1)
positive_power: Literal["charging", "discharging"]
measurement_boundary: Literal["battery_dc"] = "battery_dc"
min_soc_change_percentage: float = Field(default=20, gt=0, le=100)
max_duration_hours: float = Field(default=168, gt=0, le=744)
class CapacityModel(BaseModel):
model_config = ConfigDict(extra="forbid", allow_inf_nan=False)
class BatteryCapacityRequest(CapacityModel):
"""Caller attests both anchors are independent of the capacity being fitted.
End is the first confirmed full point when using the default end SoC.
Voltage/current anchors must be independently established for the chemistry.
"""
start: AwareDatetime
end: AwareDatetime
start_soc_percentage: float = Field(ge=0, le=100)
end_soc_percentage: float = Field(default=100, ge=0, le=100)
soc_reference: Literal["bms", "external_calibration", "voltage_current_anchor"]
store_estimate: bool = False
@model_validator(mode="after")
def ordered(self) -> "BatteryCapacityRequest":
if self.end.timestamp() <= self.start.timestamp():
raise ValueError("end must be after start.")
return self
class BatteryCapacityEstimate(CapacityModel):
battery_id: str
start: AwareDatetime
end: AwareDatetime
soc_reference: str
start_soc_percentage: float
end_soc_percentage: float
estimated_capacity_wh: float
configured_capacity_wh: float
capacity_change_percentage: float
model_end_soc_percentage_unclipped: float
model_soc_error_percentage_points: float
charge_energy_wh: float
discharge_energy_wh: float
stored_energy_change_wh: float
charging_efficiency: float
discharging_efficiency: float
power_key: str
positive_power: str
integration_method: str
coverage_seconds: float
samples_used: int
warnings: list[str]
def estimate_capacity(
request: BatteryCapacityRequest,
settings: BatteryCapacityEstimationSettings,
channel: "MeasurementChannelSettings",
samples: list[tuple[datetime, float | None, SampleQuality]],
*,
battery_id: str,
capacity_wh: float,
charging_efficiency: float,
discharging_efficiency: float,
) -> BatteryCapacityEstimate:
"""Fit C in C * delta_soc = eta_c * E_charge - E_discharge / eta_d.
No clipping at 0/100%, extrapolation, gap filling or efficiency fitting.
Linear segments are split at zero before applying directional efficiencies.
"""
left, right = request.start.timestamp(), request.end.timestamp()
if right - left > settings.max_duration_hours * 3600:
raise ValueError("Requested period exceeds capacity_estimation.max_duration_hours.")
delta_soc = (request.end_soc_percentage - request.start_soc_percentage) / 100
if abs(delta_soc) * 100 + 1e-9 < settings.min_soc_change_percentage:
raise ValueError(
"Independent SoC change is too small; full-to-full cannot identify capacity."
)
if channel.quantity != "power":
raise ValueError("Capacity estimation requires signed battery DC power in W or kW.")
max_gap_seconds = channel.max_gap_seconds
integration_method = channel.integration_method
if max_gap_seconds is None or integration_method is None:
raise ValueError("Power channels require an integration method and maximum sample gap.")
if not isfinite(capacity_wh) or capacity_wh <= 0:
raise ValueError("Configured capacity must be positive and finite.")
if not all(isfinite(e) and 0 < e <= 1 for e in (charging_efficiency, discharging_efficiency)):
raise ValueError("Battery efficiencies must be finite and in (0, 1].")
points = []
for time, value, quality in samples:
if time.tzinfo is None or time.utcoffset() is None:
raise ValueError("Explicit sample timezone required.")
points.append((time.timestamp(), value, quality))
points.sort(key=lambda row: row[0])
if any(a[0] == b[0] for a, b in zip(points, points[1:])):
raise ValueError("Duplicate power sample timestamps.")
factor = (1000 if channel.unit == "kW" else 1) * (
1 if settings.positive_power == "charging" else -1
)
charge = discharge = coverage = 0.0
cumulative_net = 0.0
minimum_net = maximum_net = 0.0
used: set[float] = set()
cursor = left
for (a, va, qa), (b, vb, qb) in zip(points, points[1:]):
lo, hi = max(left, a), min(right, b)
if hi <= lo:
continue
if lo > cursor + 1e-6 or b - a > max_gap_seconds:
raise ValueError("Incomplete power coverage or sample gap exceeds max_gap_seconds.")
if (
va is None
or vb is None
or any(isinstance(v, bool) or not isfinite(v) for v in (va, vb))
or qa.status != "measured"
or qb.status != "measured"
):
raise ValueError(
"Capacity estimation requires finite measured power samples throughout."
)
if qb.reset or qa.generation != qb.generation:
raise ValueError("Power sensor reset or generation change within the period.")
p, q = va * factor, vb * factor
if integration_method == "linear":
slope = (q - p) / (b - a)
p, q = p + slope * (lo - a), p + slope * (hi - a)
else:
q = p
duration = hi - lo
if p * q < 0:
first = duration * abs(p) / (abs(p) + abs(q))
parts = [(p * first / 7200), (q * (duration - first) / 7200)]
else:
parts = [(p + q) * duration / 7200]
charge += sum(max(0, energy) for energy in parts)
discharge += sum(max(0, -energy) for energy in parts)
for energy in parts:
cumulative_net += (
energy * charging_efficiency if energy >= 0 else energy / discharging_efficiency
)
minimum_net = min(minimum_net, cumulative_net)
maximum_net = max(maximum_net, cumulative_net)
coverage += duration
cursor = hi
used.update((a, b))
if cursor < right - 1e-6 or abs(coverage - (right - left)) > 1e-6:
raise ValueError("Incomplete power coverage; no extrapolation to the SoC anchors.")
net = charge * charging_efficiency - discharge / discharging_efficiency
capacity = net / delta_soc
if not isfinite(capacity) or capacity <= 0:
raise ValueError("Energy flow disagrees with SoC change; check polarity and anchors.")
if (
request.start_soc_percentage + minimum_net / capacity * 100 < -1e-6
or request.start_soc_percentage + maximum_net / capacity * 100 > 100 + 1e-6
):
raise ValueError(
"Fitted SoC leaves 0..100% within the period; check anchors and use the first full point."
)
model_end = request.start_soc_percentage + net / capacity_wh * 100
warnings = [
"Conditional estimate: SoC anchors, DC measurement boundary and configured efficiencies must be correct.",
"A single interval cannot independently identify both capacity and efficiencies.",
"The active capacity_wh is unchanged.",
]
if request.soc_reference == "voltage_current_anchor":
warnings.append(
"Voltage/current anchors depend on chemistry, temperature and operating conditions."
)
return BatteryCapacityEstimate(
battery_id=battery_id,
start=request.start,
end=request.end,
soc_reference=request.soc_reference,
start_soc_percentage=request.start_soc_percentage,
end_soc_percentage=request.end_soc_percentage,
estimated_capacity_wh=capacity,
configured_capacity_wh=capacity_wh,
capacity_change_percentage=(capacity / capacity_wh - 1) * 100,
model_end_soc_percentage_unclipped=model_end,
model_soc_error_percentage_points=model_end - request.end_soc_percentage,
charge_energy_wh=charge,
discharge_energy_wh=discharge,
stored_energy_change_wh=net,
charging_efficiency=charging_efficiency,
discharging_efficiency=discharging_efficiency,
power_key=settings.power_key,
positive_power=settings.positive_power,
integration_method=integration_method,
coverage_seconds=coverage,
samples_used=len(used),
warnings=warnings,
)
+190
View File
@@ -0,0 +1,190 @@
"""Quality-aware energy conversion of raw measurement samples.
No extrapolation beyond the last sample. Intervals use elapsed UTC seconds.
"""
from dataclasses import dataclass
from datetime import datetime, timezone
from math import isfinite
from typing import TYPE_CHECKING, Iterable, Literal
from akkudoktoreos.measurement.quality import SampleQuality
if TYPE_CHECKING:
from akkudoktoreos.measurement.measurement import MeasurementChannelSettings
@dataclass(frozen=True)
class EnergyInterval:
"""Energy and actual temporal support; missing energy is never zero-filled."""
start: datetime
end: datetime
energy_wh: float | None
observed_energy_wh: float | None
coverage_seconds: float
coverage_status: Literal["complete", "partial", "missing", "invalid"]
methods: tuple[str, ...]
flags: tuple[str, ...]
# Retain the support for subsequent multi-channel balance intersection.
coverage_ranges: tuple[tuple[datetime, datetime], ...]
def _timestamp(value: datetime) -> float:
if value.tzinfo is None or value.utcoffset() is None:
raise ValueError("Explicit timezone required.")
return value.timestamp()
def _date(value: float) -> datetime:
return datetime.fromtimestamp(value, timezone.utc)
def energy_intervals(
samples: Iterable[tuple[datetime, float | None]],
channel: "MeasurementChannelSettings",
start: datetime,
end: datetime,
interval_seconds: int = 900,
quality: dict[float, SampleQuality] | None = None,
) -> list[EnergyInterval]:
"""Convert finite samples, preserving gaps, resets and partial coverage.
Hold applies only between consecutive observations within max_gap_seconds.
Null/invalid observations break continuity. Meter resets invalidate that segment.
Fixed interval energy is uniformly allocated when a target cuts its source interval.
"""
left, right = _timestamp(start), _timestamp(end)
if right <= left or type(interval_seconds) is not int or interval_seconds <= 0:
raise ValueError("Require end > start and positive integer interval_seconds.")
points = sorted(((_timestamp(t), v) for t, v in samples), key=lambda point: point[0])
if any(a[0] == b[0] for a, b in zip(points, points[1:])):
raise ValueError("Duplicate sample timestamps must be resolved before conversion.")
scale = 1000 if channel.unit in ("kW", "kWh") else 1
quality = quality or {}
def sample_quality(time: float) -> SampleQuality:
return quality.get(time, SampleQuality())
def number(value: float | None) -> float | None:
if isinstance(value, bool) or not isinstance(value, (int, float)) or not isfinite(value):
return None
return float(value) * scale
# Segment: start, end, start/end power in W, method, error flag.
segments = []
if channel.quantity == "interval_energy":
duration = channel.interval_seconds
if duration is None:
raise ValueError("Interval duration is required.")
previous_end = None
for time, raw in points:
a = time if channel.timestamp_reference == "start" else time - duration
b = a + duration
if previous_end is not None and a < previous_end:
raise ValueError("Overlapping source energy intervals.")
previous_end = b
value = number(raw)
if sample_quality(time).status in ("invalid", "unavailable"):
value = None
power = value * 3600 / duration if value is not None else None
segments.append(
(a, b, power, power, "interval_energy", "invalid_sample" if power is None else None)
)
else:
for (a, raw_a), (b, raw_b) in zip(points, points[1:]):
va, vb = number(raw_a), number(raw_b)
qa, qb = sample_quality(a), sample_quality(b)
if qa.status in ("invalid", "unavailable"):
va = None
if qb.status in ("invalid", "unavailable"):
vb = None
flag = None
if channel.max_gap_seconds is not None and b - a > channel.max_gap_seconds:
flag = "gap_too_large"
elif va is None or (
vb is None
and (
channel.quantity == "cumulative_energy"
or channel.integration_method == "linear"
)
):
flag = "invalid_sample"
if channel.quantity == "cumulative_energy":
method = "meter_difference"
if va is not None and vb is not None and vb < va:
flag = "meter_reset"
if qb.reset or qa.generation != qb.generation:
flag = "meter_reset"
pa = pb = (
(vb - va) * 3600 / (b - a)
if flag is None and va is not None and vb is not None
else None
)
else:
method = "integrated_power"
pa = va
pb = vb if channel.integration_method == "linear" else va
segments.append((a, b, pa, pb, method, flag))
result = []
slot = left
segment_index = 0
while slot < right:
stop = min(slot + interval_seconds, right)
total, coverage = 0.0, 0.0
methods: set[str] = set()
flags: set[str] = set()
ranges: list[tuple[datetime, datetime]] = []
while segment_index < len(segments) and segments[segment_index][1] <= slot:
segment_index += 1
for index in range(segment_index, len(segments)):
a, b, pa, pb, method, flag = segments[index]
if a >= stop:
break
lo, hi = max(a, slot), min(b, stop)
if hi <= lo:
continue
# Quality follows the endpoints used by the integration rule.
source_time = a
if channel.quantity == "interval_energy" and channel.timestamp_reference == "end":
source_time = b
statuses = {sample_quality(source_time).status}
if channel.quantity == "cumulative_energy" or channel.integration_method == "linear":
statuses.add(sample_quality(b).status)
flags.update(s for s in statuses if s != "measured")
if flag is not None:
flags.add(flag)
continue
if pa is None or pb is None:
raise ValueError("Valid segment requires finite endpoint powers.")
p_lo = pa + (pb - pa) * (lo - a) / (b - a)
p_hi = pa + (pb - pa) * (hi - a) / (b - a)
total += (p_lo + p_hi) / 2 * (hi - lo) / 3600
coverage += hi - lo
if ranges and ranges[-1][1] == _date(lo):
ranges[-1] = (ranges[-1][0], _date(hi))
else:
ranges.append((_date(lo), _date(hi)))
methods.add(method)
if method in ("meter_difference", "interval_energy") and (lo != a or hi != b):
methods.add("allocated_energy")
complete = abs(coverage - (stop - slot)) < 1e-6
status: Literal["complete", "partial", "missing", "invalid"] = (
"complete" if complete else "partial" if coverage else "invalid" if flags else "missing"
)
result.append(
EnergyInterval(
_date(slot),
_date(stop),
total if complete else None,
total if coverage else None,
coverage,
status,
tuple(sorted(methods)),
tuple(sorted(flags)),
tuple(ranges),
)
)
slot = stop
return result
+118
View File
@@ -0,0 +1,118 @@
"""AC household balances on the intersection of actual measurement support."""
from datetime import datetime
from typing import Callable, Literal
from pydantic import BaseModel, ConfigDict, Field, model_validator
from akkudoktoreos.measurement.energy import EnergyInterval
class HouseholdInput(BaseModel):
"""A non-overlapping AC branch; polarity normalizes the sensor's sign."""
model_config = ConfigDict(extra="forbid")
key: str
branch: str = Field(min_length=1)
role: Literal["site", "grid", "pv", "battery", "inverter", "ev", "device"]
polarity: Literal[-1, 1] = 1
class HouseholdSettings(BaseModel):
"""Fixed topology, never user-supplied executable balance expressions.
Grid import, PV production and battery/inverter discharge are positive.
EV/device inputs are positive consumption, subtracted only from derived loads.
"""
model_config = ConfigDict(extra="forbid")
topology: Literal["direct", "separate_ac", "hybrid_ac"]
inputs: list[HouseholdInput] = Field(min_length=1)
@model_validator(mode="after")
def validate_topology(self) -> "HouseholdSettings":
keys = [item.key for item in self.inputs]
branches = [item.branch for item in self.inputs]
if len(keys) != len(set(keys)) or len(branches) != len(set(branches)):
raise ValueError("Duplicate measurement key or physical branch in household balance.")
roles = {item.role for item in self.inputs} - {"ev", "device"}
allowed = {
"direct": {"site"},
"separate_ac": {"grid", "pv", "battery"},
# Additional AC PV branches are independent of the hybrid's net output.
# Its DC PV/battery must not be added again as separate inputs.
"hybrid_ac": {"grid", "inverter", "pv"},
}[self.topology]
required = {"site"} if self.topology == "direct" else {"grid"}
if self.topology == "hybrid_ac":
required.add("inverter")
if not required <= roles or not roles <= allowed:
raise ValueError("Inputs do not match the selected AC topology.")
return self
def household_intervals(
settings: HouseholdSettings,
convert: Callable[[str, datetime, datetime, int], list[EnergyInterval]],
start: datetime,
end: datetime,
interval_seconds: int = 900,
) -> dict[str, list[EnergyInterval]]:
"""Integrate each source again over shared support, never prorate partial sums."""
series = {item.key: convert(item.key, start, end, interval_seconds) for item in settings.inputs}
output: dict[str, list[EnergyInterval]] = {}
for name, excluded in (
("site", {"ev", "device"}),
("household", {"device"}),
("base", set()),
):
inputs = [item for item in settings.inputs if item.role not in excluded]
result = []
for index, template in enumerate(series[inputs[0].key]):
rows = [series[item.key][index] for item in inputs]
# Partition at every coverage boundary; keep only the intersection.
boundaries = sorted({t for row in rows for pair in row.coverage_ranges for t in pair})
support = [
(a, b)
for a, b in zip(boundaries, boundaries[1:])
if all(any(lo <= a and b <= hi for lo, hi in row.coverage_ranges) for row in rows)
]
total = 0.0
methods = {method for row in rows for method in row.methods}
flags = {flag for row in rows for flag in row.flags}
for a, b in support:
for item in inputs:
# The range is at most one target interval, but may be fractional seconds.
parts = convert(item.key, a, b, interval_seconds)
value = parts[0].energy_wh
if value is None:
raise ValueError("Inconsistent measurement support during balance.")
sign = -1 if item.role in ("ev", "device") else 1
total += sign * item.polarity * value
methods.update(parts[0].methods)
flags.update(parts[0].flags)
coverage = sum((b - a).total_seconds() for a, b in support)
complete = abs(coverage - (template.end - template.start).total_seconds()) < 1e-6
if coverage and total < -1e-6:
flags.add("negative_balance")
result.append(
EnergyInterval(
template.start,
template.end,
total if complete else None,
total if coverage else None,
coverage,
"complete"
if complete
else "partial"
if coverage
else "invalid"
if flags
else "missing",
tuple(sorted(methods | {"ac_balance"})),
tuple(sorted(flags)),
tuple(support),
)
)
output[name] = result
return output
+286 -3
View File
@@ -7,17 +7,28 @@ The measurements can be added programmatically or imported from a file or JSON s
"""
import json
from bisect import bisect_left, bisect_right
from datetime import datetime, timedelta
from pathlib import Path
from typing import Any, Optional
from typing import Any, Callable, Literal, Optional
import numpy as np
from loguru import logger
from numpydantic import NDArray, Shape
from pydantic import Field, computed_field
from pydantic import Field, computed_field, model_validator
from akkudoktoreos.config.configabc import SettingsBaseModel
from akkudoktoreos.core.coreabc import SingletonMixin
from akkudoktoreos.core.dataabc import DataImportMixin, DataRecord, DataSequence
from akkudoktoreos.core.databaseabc import DatabaseTimestamp
from akkudoktoreos.measurement.batterycapacity import (
BatteryCapacityEstimate,
BatteryCapacityRequest,
estimate_capacity,
)
from akkudoktoreos.measurement.energy import EnergyInterval
from akkudoktoreos.measurement.household import HouseholdSettings, household_intervals
from akkudoktoreos.measurement.quality import MeasurementSample, SampleQuality
from akkudoktoreos.utils.datetimeutil import (
DateTime,
Duration,
@@ -26,6 +37,41 @@ from akkudoktoreos.utils.datetimeutil import (
)
class MeasurementChannelSettings(SettingsBaseModel):
"""Meaning of a raw measurement channel; no conversion is performed on storage."""
quantity: Literal["power", "cumulative_energy", "interval_energy"]
unit: Literal["W", "kW", "Wh", "kWh"]
integration_method: Optional[Literal["hold", "linear"]] = None
max_gap_seconds: Optional[int] = Field(default=None, gt=0, strict=True)
interval_seconds: Optional[int] = Field(default=None, gt=0, strict=True)
timestamp_reference: Optional[Literal["start", "end"]] = None
@model_validator(mode="after")
def validate_semantics(self) -> "MeasurementChannelSettings":
"""Reject ambiguous units and time semantics before accepting a channel."""
if self.quantity == "power":
if self.unit not in ("W", "kW"):
raise ValueError("Power channels require W or kW.")
if self.integration_method is None or self.max_gap_seconds is None:
raise ValueError("Power channels require integration_method and max_gap_seconds.")
else:
if self.unit not in ("Wh", "kWh"):
raise ValueError("Energy channels require Wh or kWh.")
if self.integration_method is not None:
raise ValueError("integration_method is only applicable to power channels.")
if self.quantity == "interval_energy":
if self.interval_seconds is None or self.timestamp_reference is None:
raise ValueError(
"Interval energy requires interval_seconds and timestamp_reference."
)
if self.max_gap_seconds is not None:
raise ValueError("Interval energy uses explicit intervals, not max_gap_seconds.")
elif self.interval_seconds is not None or self.timestamp_reference is not None:
raise ValueError("Interval metadata is only applicable to interval_energy channels.")
return self
class MeasurementCommonSettings(SettingsBaseModel):
"""Measurement Configuration."""
@@ -38,6 +84,47 @@ class MeasurementCommonSettings(SettingsBaseModel):
},
)
channels: dict[str, MeasurementChannelSettings] = Field(
default_factory=dict,
json_schema_extra={
"description": "Typed raw measurement channels keyed by measurement key."
},
)
household: Optional[HouseholdSettings] = Field(
default=None,
json_schema_extra={
"description": "Optional household energy balance definition.",
"examples": [None],
},
)
energy_context_seconds: int = Field(default=86400, gt=0, le=604800, strict=True)
@model_validator(mode="after")
def validate_channels(self) -> "MeasurementCommonSettings":
"""Preserve the kWh meter contract of legacy keys and avoid record collisions."""
for key, channel in self.channels.items():
if (
not key
or key != key.strip()
or key.startswith("_")
or hasattr(DataRecord, key)
or key in DataRecord.model_fields
or key == "sample_quality"
):
raise ValueError(f"Invalid or reserved measurement channel key: {key!r}")
for name in type(self).model_fields:
if name.endswith("_emr_keys") and key in (getattr(self, name) or []):
if channel.quantity != "cumulative_energy" or channel.unit != "kWh":
raise ValueError(
f"Legacy meter key {key!r} must remain cumulative_energy in kWh."
)
if self.household is not None:
for item in self.household.inputs:
if item.key not in self.keys:
raise ValueError(f"Unknown household measurement key: {item.key!r}")
return self
load_emr_keys: Optional[list[str]] = Field(
default=None,
json_schema_extra={
@@ -75,7 +162,7 @@ class MeasurementCommonSettings(SettingsBaseModel):
@property
def keys(self) -> list[str]:
"""The keys of the measurements that can be stored."""
key_list = []
key_list = list(self.channels)
for key in self.__class__.model_fields.keys():
if key.endswith("_keys") and (value := getattr(self, key)):
key_list.extend(value)
@@ -85,6 +172,18 @@ class MeasurementCommonSettings(SettingsBaseModel):
class MeasurementDataRecord(DataRecord):
"""Represents a measurement data record containing various measurements at a specific datetime."""
sample_quality: dict[str, SampleQuality] = Field(default_factory=dict)
@classmethod
def record_keys(cls) -> list[str]:
"""Quality is stored alongside values, not exposed as a numeric channel."""
return [key for key in super().record_keys() if key != "sample_quality"]
@classmethod
def record_keys_writable(cls) -> list[str]:
"""Only the typed sample path writes quality, never numeric import paths."""
return [key for key in super().record_keys_writable() if key != "sample_quality"]
@classmethod
def configured_data_keys(cls) -> Optional[list[str]]:
"""Return the keys for the configured field like data."""
@@ -150,6 +249,190 @@ class Measurement(SingletonMixin, DataImportMixin, DataSequence[MeasurementDataR
# Return ceiling of division to include partial intervals
return int(np.ceil(diff_seconds / interval_seconds))
async def import_samples(self, samples: list[MeasurementSample]) -> None:
"""Validate the whole batch before replacing samples; omitted keys stay untouched."""
for sample in samples:
self._energy_channel(sample.key)
for sample in samples:
dt = to_datetime(sample.date_time)
await self.update_value(dt, sample.key, sample.value)
record = await self.db_get_record(DatabaseTimestamp.from_datetime(dt))
if not isinstance(record, MeasurementDataRecord):
raise RuntimeError("Measurement sample was not stored.")
record.sample_quality[sample.key] = sample.quality.model_copy(deep=True)
await self.db_mark_dirty_record(record)
async def insert_by_datetime(self, record: MeasurementDataRecord) -> None:
"""Merge quality by channel as well as the ordinary measurement fields."""
await super().insert_by_datetime(record)
if (
isinstance(record, MeasurementDataRecord)
and record.sample_quality
and record.date_time is not None
):
stored = await self.db_get_record(DatabaseTimestamp.from_datetime(record.date_time))
if not isinstance(stored, MeasurementDataRecord):
raise RuntimeError("Measurement record was not stored.")
stored.sample_quality = stored.sample_quality | record.sample_quality
await self.db_mark_dirty_record(stored)
def _energy_channel(self, key: str) -> MeasurementChannelSettings:
channel = self.config.measurement.channels.get(key)
if channel is not None:
return channel
for name in type(self.config.measurement).model_fields:
if name.endswith("_emr_keys") and key in (getattr(self.config.measurement, name) or []):
return MeasurementChannelSettings(quantity="cumulative_energy", unit="kWh")
raise ValueError(f"No energy channel definition for {key!r}.")
async def _energy_converter(
self,
keys: list[str],
start: datetime,
end: datetime,
) -> Callable[[str, datetime, datetime, int], list[EnergyInterval]]:
"""Load a bounded window once, including configured boundary context."""
from akkudoktoreos.measurement.energy import energy_intervals
if any(dt.tzinfo is None or dt.utcoffset() is None for dt in (start, end)):
raise ValueError("Explicit timezone required.")
seconds = end.timestamp() - start.timestamp()
if not 0 < seconds <= 31 * 86400:
raise ValueError("Energy queries require a positive range of at most 31 days.")
channels = {key: self._energy_channel(key) for key in keys}
context = self.config.measurement.energy_context_seconds
records = [
record
async for record in self.db_iterate_records(
DatabaseTimestamp.from_datetime(to_datetime(start) - timedelta(seconds=context)),
DatabaseTimestamp.from_datetime(
to_datetime(end) + timedelta(seconds=context, microseconds=1)
),
)
]
samples = {
key: [
(record.date_time, record.configured_data[key])
for record in records
if key in record.configured_data and record.date_time is not None
]
for key in keys
}
quality = {
key: {
record.date_time.timestamp(): record.sample_quality[key]
for record in records
if key in record.sample_quality and record.date_time is not None
}
for key in keys
}
timestamps = {key: [dt.timestamp() for dt, _ in values] for key, values in samples.items()}
for key, channel in channels.items():
duration = channel.interval_seconds
if (
channel.quantity == "interval_energy"
and duration is not None
and any(b - a < duration for a, b in zip(timestamps[key], timestamps[key][1:]))
):
raise ValueError(f"Overlapping source energy intervals for {key!r}.")
def convert(
key: str, left: datetime, right: datetime, interval_seconds: int
) -> list[EnergyInterval]:
if type(interval_seconds) is not int or interval_seconds <= 0:
raise ValueError("Positive integer interval_seconds required.")
if (right.timestamp() - left.timestamp()) / interval_seconds > 10000:
raise ValueError("At most 10000 output intervals per query.")
lo = max(0, bisect_left(timestamps[key], left.timestamp()) - 1)
hi = bisect_right(timestamps[key], right.timestamp()) + 1
return energy_intervals(
samples[key][lo:hi], channels[key], left, right, interval_seconds, quality[key]
)
return convert
async def household_intervals(
self, start_datetime: datetime, end_datetime: datetime, interval_seconds: int = 900
) -> dict[str, list[EnergyInterval]]:
"""Return site, household without EV, and base without configured devices."""
settings = self.config.measurement.household
if settings is None:
raise ValueError("No household balance configured.")
# Revalidate references even after an in-place configuration mutation.
settings = HouseholdSettings.model_validate(settings.model_dump())
convert = await self._energy_converter(
[item.key for item in settings.inputs], start_datetime, end_datetime
)
return household_intervals(
settings, convert, start_datetime, end_datetime, interval_seconds
)
async def energy_intervals(
self,
key: str,
start_datetime: datetime,
end_datetime: datetime,
interval_seconds: int = 900,
) -> list[EnergyInterval]:
"""Convert a typed channel without changing the legacy kWh calculation.
Read only explicitly stored values for this key: another channel's timestamp
must not introduce a synthetic outage. Explicit null values remain barriers.
"""
convert = await self._energy_converter([key], start_datetime, end_datetime)
return convert(key, start_datetime, end_datetime, interval_seconds)
async def estimate_battery_capacity(
self, battery_id: str, request: BatteryCapacityRequest
) -> BatteryCapacityEstimate:
"""Read signed DC samples without mutating raw data or the active capacity."""
batteries = [
b for b in (self.config.devices.batteries or {}).values() if b.device_id == battery_id
]
if len(batteries) != 1:
raise ValueError("Require exactly one configured battery with this device_id.")
battery = batteries[0]
settings = battery.capacity_estimation
if settings is None:
raise ValueError("Configure devices.batteries[device_id].capacity_estimation first.")
if request.end.timestamp() - request.start.timestamp() > settings.max_duration_hours * 3600:
raise ValueError("Requested period exceeds capacity_estimation.max_duration_hours.")
channel = self._energy_channel(settings.power_key)
if channel.quantity != "power":
raise ValueError("Capacity estimation requires a battery DC power channel.")
context = channel.max_gap_seconds
if context is None:
raise ValueError("Capacity estimation requires a bounded power channel gap.")
samples: list[tuple[datetime, Optional[float], SampleQuality]] = []
async for record in self.db_iterate_records(
DatabaseTimestamp.from_datetime(
to_datetime(request.start) - timedelta(seconds=context)
),
DatabaseTimestamp.from_datetime(
to_datetime(request.end) + timedelta(seconds=context, microseconds=1)
),
):
if record.date_time is not None and settings.power_key in record.configured_data:
samples.append(
(
record.date_time,
record.configured_data[settings.power_key],
record.sample_quality.get(settings.power_key, SampleQuality()),
)
)
if len(samples) > 250000:
raise ValueError("More than 250000 power samples; request a shorter period.")
return estimate_capacity(
request,
settings,
channel,
samples,
battery_id=battery.device_id,
capacity_wh=battery.capacity_wh,
charging_efficiency=battery.charging_efficiency,
discharging_efficiency=battery.discharging_efficiency,
)
async def _energy_from_meter_readings(
self,
key: str,
+24
View File
@@ -0,0 +1,24 @@
"""Optional quality information on raw measurement samples."""
from typing import Literal
from pydantic import AwareDatetime, BaseModel, ConfigDict, Field
class SampleQuality(BaseModel):
"""Reset marks the first reading after a reset; generation identifies a meter."""
model_config = ConfigDict(extra="forbid")
status: Literal["measured", "estimated", "invalid", "unavailable"] = "measured"
reset: bool = False
generation: str | None = Field(default=None, max_length=128)
class MeasurementSample(BaseModel):
"""Complete replacement of one key/timestamp, including its quality."""
model_config = ConfigDict(extra="forbid", allow_inf_nan=False)
date_time: AwareDatetime
key: str
value: float | None = Field(strict=True)
quality: SampleQuality = Field(default_factory=SampleQuality)
+3
View File
@@ -80,6 +80,7 @@ from akkudoktoreos.server.rest.error import (
create_error_page,
register_problem_handlers,
)
from akkudoktoreos.server.rest.measurement import router as measurement_router
from akkudoktoreos.server.rest.starteosdash import supervise_eosdash
from akkudoktoreos.server.retentionmanager import RetentionManager
from akkudoktoreos.server.server import (
@@ -261,6 +262,8 @@ The genetic optimization API fields were renamed from German to English. For bac
# Application generic exception handling
# ----------------------
app.include_router(measurement_router)
register_problem_handlers(app)
@@ -0,0 +1,114 @@
"""Typed samples and derived energy within the existing measurement API."""
from typing import Annotated
from fastapi import APIRouter, Body, HTTPException, Query
from pydantic import AwareDatetime
from akkudoktoreos.core.coreabc import get_config, get_measurement
from akkudoktoreos.core.databaseabc import DatabaseTimestamp
from akkudoktoreos.measurement.batterycapacity import (
BatteryCapacityEstimate,
BatteryCapacityRequest,
)
from akkudoktoreos.measurement.energy import EnergyInterval
from akkudoktoreos.measurement.quality import MeasurementSample, SampleQuality
from akkudoktoreos.utils.datetimeutil import to_datetime
router = APIRouter(prefix="/v1/measurement", tags=["measurement"])
@router.post("/battery-capacity/{battery_id}", response_model=BatteryCapacityEstimate)
async def post_battery_capacity(
battery_id: str, request: BatteryCapacityRequest
) -> BatteryCapacityEstimate:
"""Estimate capacity from independent SoC anchors and configured DC power.
store_estimate writes the separate capacity_estimate config field in memory.
Persistence follows the regular EOS configuration save mechanism.
The active capacity_wh and raw measurements are never changed here.
"""
try:
estimate = await get_measurement().estimate_battery_capacity(battery_id, request)
if request.store_estimate:
batteries = [
b
for b in (get_config().devices.batteries or {}).values()
if b.device_id == battery_id
]
if len(batteries) != 1:
raise ValueError("Battery configuration changed during estimation; retry.")
get_config().set_nested_value(
f"devices/batteries/{battery_id}/capacity_estimate", estimate
)
return estimate
except ValueError as exc:
raise HTTPException(422, str(exc)) from exc
@router.put("/samples")
async def put_samples(
samples: Annotated[list[MeasurementSample], Body(max_length=10000)],
) -> dict[str, int]:
"""Upsert raw values and their quality; legacy value/series payloads remain valid."""
try:
await get_measurement().import_samples(samples)
except ValueError as exc:
raise HTTPException(422, str(exc)) from exc
return {"updated": len(samples)}
@router.get("/samples", response_model=list[MeasurementSample])
async def get_samples(
key: str, start: AwareDatetime, end: AwareDatetime
) -> list[MeasurementSample]:
"""Read raw samples including quality, in a bounded half-open range."""
measurement = get_measurement()
try:
measurement._energy_channel(key)
if not 0 < end.timestamp() - start.timestamp() <= 31 * 86400:
raise ValueError("Require a positive range of at most 31 days.")
result = []
async for record in measurement.db_iterate_records(
DatabaseTimestamp.from_datetime(to_datetime(start)),
DatabaseTimestamp.from_datetime(to_datetime(end)),
):
if key in record.configured_data and record.date_time is not None:
result.append(
MeasurementSample(
date_time=record.date_time,
key=key,
value=record.configured_data[key],
quality=record.sample_quality.get(key, SampleQuality()),
)
)
if len(result) > 10000:
raise ValueError("More than 10000 samples; request a shorter range.")
return result
except ValueError as exc:
raise HTTPException(422, str(exc)) from exc
@router.get("/energy", response_model=list[EnergyInterval])
async def get_energy(
key: str,
start: AwareDatetime,
end: AwareDatetime,
interval_seconds: Annotated[int, Query(gt=0)] = 900,
) -> list[EnergyInterval]:
"""Energy in Wh, with temporal coverage and quality; no missing-to-zero filling."""
try:
return await get_measurement().energy_intervals(key, start, end, interval_seconds)
except ValueError as exc:
raise HTTPException(422, str(exc)) from exc
@router.get("/household", response_model=dict[str, list[EnergyInterval]])
async def get_household(
start: AwareDatetime, end: AwareDatetime, interval_seconds: Annotated[int, Query(gt=0)] = 900
) -> dict[str, list[EnergyInterval]]:
"""Site, household without EV, and base without separately measured devices."""
try:
return await get_measurement().household_intervals(start, end, interval_seconds)
except ValueError as exc:
raise HTTPException(422, str(exc)) from exc
+246
View File
@@ -0,0 +1,246 @@
import pytest_asyncio
"""Capacity fits must preserve energy direction, coverage and independent anchors."""
# ruff: noqa: S101
from datetime import datetime, timedelta, timezone
import pytest
from pydantic import ValidationError
from akkudoktoreos.measurement.batterycapacity import (
BatteryCapacityEstimationSettings,
BatteryCapacityRequest,
estimate_capacity,
)
from akkudoktoreos.measurement.measurement import MeasurementChannelSettings
from akkudoktoreos.measurement.quality import SampleQuality
START = datetime(2026, 9, 10, tzinfo=timezone.utc)
def fit(
points,
*,
start_soc=20,
end_soc=100,
efficiency=1,
method="hold",
polarity="charging",
unit="W",
end_seconds=3600,
quality=None,
max_gap=3600,
):
request = BatteryCapacityRequest.model_validate(dict(
start=START,
end=START + timedelta(seconds=end_seconds),
start_soc_percentage=start_soc,
end_soc_percentage=end_soc,
soc_reference="voltage_current_anchor",
))
settings = BatteryCapacityEstimationSettings(power_key="dc", positive_power=polarity)
channel = MeasurementChannelSettings(
quantity="power",
unit=unit,
integration_method=method,
max_gap_seconds=max_gap,
)
return estimate_capacity(
request,
settings,
channel,
[
(START + timedelta(seconds=t), v, (quality or {}).get(t, SampleQuality()))
for t, v in points
],
battery_id="battery",
capacity_wh=12000,
charging_efficiency=efficiency,
discharging_efficiency=efficiency,
)
def test_charge_fit_and_unclipped_model_error():
result = fit([(0, 8000), (3600, 8000)])
assert result.estimated_capacity_wh == pytest.approx(10000)
assert result.configured_capacity_wh == 12000
assert result.model_end_soc_percentage_unclipped == pytest.approx(86.6666667)
assert result.model_soc_error_percentage_points == pytest.approx(-13.3333333)
assert result.charge_energy_wh == 8000
assert result.coverage_seconds == 3600
def test_efficiency_is_applied_once_on_dc_boundary():
result = fit([(0, 10000), (3600, 10000)], efficiency=0.8)
assert result.estimated_capacity_wh == pytest.approx(10000)
assert result.stored_energy_change_wh == 8000
def test_discharge_fit_and_reversed_sensor_sign():
result = fit(
[(0, 6400), (3600, 6400)], start_soc=100, end_soc=20, polarity="discharging", efficiency=0.8
)
assert result.discharge_energy_wh == 6400
assert result.estimated_capacity_wh == pytest.approx(10000)
def test_linear_zero_crossing_is_split_before_losses():
result = fit([(0, -4000), (3600, 12000)], method="linear", efficiency=0.8)
assert result.charge_energy_wh == pytest.approx(4500)
assert result.discharge_energy_wh == pytest.approx(500)
assert result.stored_energy_change_wh == pytest.approx(4500 * 0.8 - 500 / 0.8)
def test_kw_and_clipped_boundary_interpolation():
result = fit([(-3600, 0), (3600, 16)], method="linear", unit="kW", max_gap=7200)
assert result.charge_energy_wh == pytest.approx(12000)
# Deliberately do not saturate the old model at 100%.
assert result.model_end_soc_percentage_unclipped == pytest.approx(120)
@pytest.mark.parametrize(
"points", [[], [(0, 8000)], [(60, 8000), (3600, 8000)], [(0, 8000), (3500, 8000)]]
)
def test_missing_coverage_is_not_extrapolated(points):
with pytest.raises(ValueError, match="coverage"):
fit(points)
def test_gap_is_rejected():
with pytest.raises(ValueError, match="gap"):
fit([(0, 8000), (3600, 8000)], max_gap=300)
@pytest.mark.parametrize("value", [None, float("nan"), float("inf"), True])
def test_bad_power_is_rejected(value):
with pytest.raises(ValueError, match="finite measured"):
fit([(0, value), (3600, 8000)])
@pytest.mark.parametrize("status", ["estimated", "invalid", "unavailable"])
def test_nonmeasured_quality_is_rejected(status):
with pytest.raises(ValueError, match="finite measured"):
fit([(0, 8000), (3600, 8000)], quality={0: SampleQuality(status=status)})
def test_reset_and_sensor_change_are_rejected():
for quality in (SampleQuality(reset=True), SampleQuality(generation="replacement")):
with pytest.raises(ValueError, match="reset or generation"):
fit([(0, 8000), (3600, 8000)], quality={3600: quality})
@pytest.mark.parametrize("start_soc", [100, 99, 85])
def test_full_to_full_and_small_soc_span_do_not_produce_estimates(start_soc):
with pytest.raises(ValueError, match="SoC change is too small"):
fit([(0, 8000), (3600, 8000)], start_soc=start_soc)
def test_wrong_polarity_is_rejected():
with pytest.raises(ValueError, match="disagrees"):
fit([(0, -8000), (3600, -8000)])
def test_hidden_saturation_cannot_be_fixed_by_end_point_fitting():
with pytest.raises(ValueError, match="Fitted SoC leaves"):
fit([(0, 24000), (1800, -8000), (3600, -8000)])
def test_model_soc_is_not_an_accepted_reference():
with pytest.raises(ValidationError):
BatteryCapacityRequest.model_validate(dict(
start=START,
end=START + timedelta(hours=1),
start_soc_percentage=20,
soc_reference="calculated_soc",
))
@pytest_asyncio.fixture
async def database_case(config_eos):
from akkudoktoreos.core.coreabc import get_measurement
from akkudoktoreos.measurement.quality import MeasurementSample
get_measurement()._db_reset_state()
config_eos.merge_settings_from_dict(
{
"devices": {
"batteries": {
"battery": {
"device_id": "battery",
"capacity_wh": 12000,
"charging_efficiency": 1,
"discharging_efficiency": 1,
"capacity_estimation": {
"power_key": "battery_dc",
"positive_power": "charging",
},
}
}
},
"measurement": {
"channels": {
"battery_dc": {
"quantity": "power",
"unit": "W",
"integration_method": "hold",
"max_gap_seconds": 3600,
}
}
},
}
)
(await get_measurement().import_samples(
[
MeasurementSample(
date_time=START + timedelta(seconds=t), key="battery_dc", value=8000.0
)
for t in (0, 1800, 3600)
]
))
try:
yield config_eos
finally:
get_measurement()._db_reset_state()
def test_http_reads_database_and_stores_only_explicit_estimate(database_case):
from fastapi.testclient import TestClient
from akkudoktoreos.server.eos import app
client = TestClient(app)
body = {
"start": START.isoformat(),
"end": (START + timedelta(hours=1)).isoformat(),
"start_soc_percentage": 20,
"soc_reference": "voltage_current_anchor",
}
response = client.post("/v1/measurement/battery-capacity/battery", json=body)
assert response.status_code == 200, response.text
assert response.json()["estimated_capacity_wh"] == pytest.approx(10000)
battery = database_case.devices.batteries["battery"]
assert battery.capacity_estimate is None
assert battery.capacity_wh == 12000
body["store_estimate"] = True
response = client.post("/v1/measurement/battery-capacity/battery", json=body)
assert response.status_code == 200, response.text
battery = database_case.devices.batteries["battery"]
assert battery.capacity_estimate is not None
assert battery.capacity_estimate.estimated_capacity_wh == pytest.approx(10000)
assert battery.capacity_wh == 12000
database_case.merge_settings_from_dict({"optimization": {"genetic": {"individuals": 100}}})
battery = database_case.devices.batteries["battery"]
assert battery.capacity_estimate is not None
assert battery.capacity_estimate.estimated_capacity_wh == pytest.approx(10000)
assert battery.capacity_wh == 12000
# The estimate survives normal config serialization without becoming capacity_wh.
data = database_case.to_config_json()
assert '"estimated_capacity_wh": 10000.0' in data
assert '"capacity_wh": 12000' in data
previous = battery.capacity_estimate
body["start_soc_percentage"] = 100
assert client.post("/v1/measurement/battery-capacity/battery", json=body).status_code == 422
assert battery.capacity_estimate is previous
+106
View File
@@ -0,0 +1,106 @@
from unittest.mock import AsyncMock
"""Contracts for typed channels sharing the existing measurement storage."""
# ruff: noqa: S101
import pytest
from pydantic import ValidationError
from akkudoktoreos.core.coreabc import get_measurement
from akkudoktoreos.measurement.measurement import (
MeasurementChannelSettings,
MeasurementCommonSettings,
MeasurementDataRecord,
)
from akkudoktoreos.utils.datetimeutil import to_datetime
CHANNELS = {
"house_power": dict(quantity="power", unit="W", integration_method="hold", max_gap_seconds=120),
"house_meter": dict(quantity="cumulative_energy", unit="kWh"),
"house_interval": dict(
quantity="interval_energy", unit="Wh", interval_seconds=900, timestamp_reference="start"
),
}
@pytest.mark.parametrize(
"definition",
[
dict(quantity="power", unit="kWh", integration_method="hold", max_gap_seconds=60),
dict(quantity="power", unit="W"),
dict(quantity="power", unit="W", integration_method="hold", max_gap_seconds=0),
dict(quantity="cumulative_energy", unit="W"),
dict(quantity="cumulative_energy", unit="kWh", timestamp_reference="end"),
dict(quantity="interval_energy", unit="Wh", interval_seconds=900),
dict(
quantity="interval_energy", unit="Wh", interval_seconds=True, timestamp_reference="end"
),
],
)
def test_reject_ambiguous_channel(definition):
with pytest.raises(ValidationError):
MeasurementChannelSettings(**definition)
@pytest.mark.parametrize(
"legacy_field",
["load_emr_keys", "grid_import_emr_keys", "grid_export_emr_keys", "pv_production_emr_keys"],
)
def test_legacy_keys_keep_meter_semantics(legacy_field):
settings = MeasurementCommonSettings.model_validate(dict(**{legacy_field: ["legacy"]}, channels=CHANNELS))
assert settings.keys == sorted(["legacy", *CHANNELS])
compatible = MeasurementCommonSettings.model_validate(dict(
**{legacy_field: ["legacy"]}, channels={"legacy": CHANNELS["house_meter"]}
))
assert compatible.keys == ["legacy"]
with pytest.raises(ValidationError, match="must remain"):
MeasurementCommonSettings.model_validate(dict(
**{legacy_field: ["legacy"]}, channels={"legacy": CHANNELS["house_power"]}
))
@pytest.mark.parametrize("key", ["", " x", "date_time", "configured_data", "keys", "_private"])
def test_reject_reserved_key(key):
with pytest.raises(ValidationError):
MeasurementCommonSettings.model_validate(dict(channels={key: CHANNELS["house_meter"]}))
@pytest.mark.asyncio
async def test_existing_import_and_file_reload(config_eos, tmp_path, monkeypatch):
"""All three quantities retain their raw values through the existing file path."""
from akkudoktoreos.core.dataabc import DataSequence
measurement = get_measurement()
previous_settings = config_eos.measurement
previous_records = measurement.records
previous_folder = config_eos.general.data_folder_path
try:
config_eos.measurement = MeasurementCommonSettings.model_validate(dict(channels=CHANNELS))
config_eos.general.data_folder_path = tmp_path
measurement._db_reset_state()
values = dict(house_power=800.0, house_meter=12345.6, house_interval=200.0)
for key, value in values.items():
(await measurement.update_value(to_datetime("2026-09-10T18:00:00Z"), key, value))
assert set(values).issubset(measurement.record_keys)
monkeypatch.setattr(DataSequence, "save", AsyncMock(return_value=False))
monkeypatch.setattr(DataSequence, "load", AsyncMock(return_value=False))
assert (await measurement.save())
measurement._db_reset_state()
assert (await measurement.load())
assert len(measurement.records) == 1
for key, value in values.items():
assert measurement.records[0][key] == value
restored = MeasurementDataRecord.model_validate_json(
measurement.records[0].model_dump_json()
)
assert restored.configured_data == values
settings = MeasurementCommonSettings.model_validate_json(
config_eos.measurement.model_dump_json()
)
assert settings.model_dump() == config_eos.measurement.model_dump()
finally:
measurement._db_reset_state()
measurement.records = previous_records
config_eos.measurement = previous_settings
config_eos.general.data_folder_path = previous_folder
+135
View File
@@ -0,0 +1,135 @@
"""Physical and temporal contracts for interval energy."""
# ruff: noqa: S101
from datetime import datetime, timedelta, timezone
from typing import Any
from zoneinfo import ZoneInfo
import pytest
from akkudoktoreos.measurement.energy import energy_intervals
from akkudoktoreos.measurement.measurement import MeasurementChannelSettings
from akkudoktoreos.utils.datetimeutil import to_datetime
START = datetime(2026, 9, 10, tzinfo=timezone.utc)
def channel(quantity="power", **kwargs):
defaults: dict[str, dict[str, Any]] = {
"power": dict(unit="W", integration_method="hold", max_gap_seconds=900),
"cumulative_energy": dict(unit="kWh"),
"interval_energy": dict(unit="Wh", interval_seconds=900, timestamp_reference="start"),
}
return MeasurementChannelSettings(quantity=quantity, **(defaults[quantity] | kwargs))
def convert(points, config=None, seconds=900):
return energy_intervals(
[(START + timedelta(seconds=t), v) for t, v in points],
config or channel(),
START,
START + timedelta(seconds=seconds),
)
@pytest.mark.parametrize(
"config,points",
[
(channel(), [(0, 800), (900, 800)]),
(channel(unit="kW"), [(0, 0.8), (900, 0.8)]),
(channel("cumulative_energy"), [(0, 10), (900, 10.2)]),
(channel("interval_energy"), [(0, 200)]),
(channel("interval_energy", timestamp_reference="end"), [(900, 200)]),
],
)
def test_equivalent_measurements(config, points):
result = convert(points, config)[0]
assert result.energy_wh == pytest.approx(200)
assert result.coverage_seconds == 900
assert result.coverage_status == "complete"
def test_time_weighting_and_linear_interpolation():
assert convert([(0, 0), (600, 1200), (900, 1200)])[0].energy_wh == 100
assert convert([(0, 0), (900, 1600)], channel(integration_method="linear"))[0].energy_wh == 200
def test_gap_and_no_extrapolation():
result = convert([(0, 800), (300, 800)])[0]
assert result.energy_wh is None
assert result.observed_energy_wh == pytest.approx(800 / 12)
assert result.coverage_status == "partial"
assert convert([(0, 800), (900, 800)], channel(max_gap_seconds=60))[0].energy_wh is None
assert convert([])[0].coverage_status == "missing"
def test_null_breaks_hold_at_outage():
result = convert([(0, 800), (300, None), (600, 800), (900, 800)])[0]
assert result.coverage_seconds == 600
assert result.coverage_status == "partial"
assert len(result.coverage_ranges) == 2
def test_reset_does_not_create_negative_consumption():
result = convert([(0, 10), (450, 0), (900, 0.1)], channel("cumulative_energy"))[0]
assert result.energy_wh is None
assert result.observed_energy_wh == pytest.approx(100)
assert "meter_reset" in result.flags
def test_hour_allocation_conserves_energy_and_is_labelled():
result = convert([(0, 1000)], channel("interval_energy", interval_seconds=3600), 3600)
assert all(r.energy_wh is not None for r in result)
assert sum(r.energy_wh for r in result if r.energy_wh is not None) == 1000
assert all("allocated_energy" in r.methods for r in result)
@pytest.mark.parametrize(
"quantity,points",
[
("power", [(0, 1), (0, 2)]),
("interval_energy", [(0, 100), (450, 100)]),
],
)
def test_ambiguous_time_support_rejected(quantity, points):
with pytest.raises(ValueError):
convert(points, channel(quantity))
def test_nan_and_signed_power():
assert convert([(0, float("nan")), (900, 800)])[0].energy_wh is None
assert convert([(0, -800), (900, -800)])[0].energy_wh == -200
@pytest.mark.parametrize("month,day,hours", [(3, 29, 23), (10, 25, 25)])
def test_dst_calendar_day(month, day, hours):
start = datetime(2026, month, day, tzinfo=ZoneInfo("Europe/Berlin"))
end = start + timedelta(days=1)
result = energy_intervals([(start, 0), (end, hours)], channel("cumulative_energy"), start, end)
assert len(result) == hours * 4
assert all(r.energy_wh is not None for r in result)
assert sum(r.energy_wh for r in result if r.energy_wh is not None) == pytest.approx(hours * 1000)
@pytest.mark.asyncio
async def test_measurement_wrapper_preserves_asynchronous_channels(config_eos):
from akkudoktoreos.core.coreabc import get_measurement
from akkudoktoreos.measurement.measurement import MeasurementCommonSettings
measurement = get_measurement()
previous, records = config_eos.measurement, measurement.records
try:
config_eos.measurement = MeasurementCommonSettings(
channels={"p": channel(), "other": channel()}
)
measurement._db_reset_state()
(await measurement.update_value(to_datetime(START), "p", 800))
(await measurement.update_value(to_datetime(START + timedelta(seconds=450)), "other", 1))
(await measurement.update_value(to_datetime(START + timedelta(seconds=900)), "p", 800))
result = (await measurement.energy_intervals("p", START, START + timedelta(seconds=900)))
assert result[0].energy_wh == 200
finally:
measurement._db_reset_state()
measurement.records = records
config_eos.measurement = previous
+366
View File
@@ -0,0 +1,366 @@
"""API, persistence and physical balance contracts for typed measurements."""
# ruff: noqa: S101
from datetime import datetime, timedelta, timezone
import pytest
from fastapi import FastAPI
from fastapi.testclient import TestClient
from pydantic import ValidationError
from akkudoktoreos.core.coreabc import get_database, get_measurement
from akkudoktoreos.measurement.measurement import MeasurementCommonSettings
from akkudoktoreos.measurement.quality import MeasurementSample
from akkudoktoreos.server.rest.measurement import router
START = datetime(2026, 9, 10, tzinfo=timezone.utc)
END = START + timedelta(seconds=900)
POWER = dict(quantity="power", unit="W", integration_method="hold", max_gap_seconds=900)
@pytest.fixture
def setup_measurement(config_eos):
measurement = get_measurement()
settings, records = config_eos.measurement, measurement.records
measurement._db_reset_state()
def configure(topology="direct", inputs=None, channels=None):
inputs = inputs or [dict(key="site", branch="house", role="site")]
config_eos.measurement = MeasurementCommonSettings.model_validate(dict(
channels=channels or {item["key"]: POWER for item in inputs},
household=dict(topology=topology, inputs=inputs),
))
return measurement
yield configure
measurement._db_reset_state()
measurement.records = records
config_eos.measurement = settings
async def write(measurement, key, points):
(await measurement.import_samples(
[
MeasurementSample(
date_time=START + timedelta(seconds=t),
key=key,
value=value,
quality=quality if quality else {},
)
for t, value, quality in points
]
))
async def constant(measurement, key, watts):
(await write(measurement, key, [(0, watts, None), (900, watts, None)]))
@pytest.mark.asyncio
@pytest.mark.parametrize(
"topology,values",
[
("direct", {"site": 4000}),
("separate_ac", {"grid": 1000, "pv": 2000, "battery": 1000}),
("hybrid_ac", {"grid": 1000, "inverter": 3000}),
],
)
async def test_topologies_and_device_subtraction(setup_measurement, topology, values):
values = values | {"ev": 2000, "device": 800}
m = setup_measurement(topology, [dict(key=k, branch=k, role=k) for k in values])
for key, value in values.items():
(await constant(m, key, value))
rows = (await m.household_intervals(START, END))
assert rows["site"][0].energy_wh == 1000
assert rows["household"][0].energy_wh == 500
assert rows["base"][0].energy_wh == 300
@pytest.mark.asyncio
async def test_export_charge_and_polarity(setup_measurement):
m = setup_measurement(
"separate_ac",
[
dict(key="grid", branch="grid", role="grid"),
dict(key="solar", branch="solar", role="pv"),
dict(key="charge", branch="battery", role="battery", polarity=-1),
],
)
for key, watts in dict(grid=-1000, solar=4000, charge=1000).items():
(await constant(m, key, watts))
assert (await m.household_intervals(START, END))["site"][0].energy_wh == 500
@pytest.mark.asyncio
@pytest.mark.parametrize("grid,inverter", [(-6000, 7200), (4200, -3000)])
async def test_hybrid_with_independent_ac_pv_and_ev(setup_measurement, grid, inverter):
inputs = [dict(key=k, branch=k, role=k) for k in ("grid", "inverter", "pv", "ev")]
m = setup_measurement("hybrid_ac", inputs)
for key, watts in dict(grid=grid, inverter=inverter, pv=400, ev=600).items():
(await constant(m, key, watts))
rows = (await m.household_intervals(START, END))
assert rows["site"][0].energy_wh == 400
assert rows["household"][0].energy_wh == 250
@pytest.mark.asyncio
async def test_hybrid_missing_ac_pv_is_not_zero(setup_measurement):
m = setup_measurement(
"hybrid_ac", [dict(key=k, branch=k, role=k) for k in ("grid", "inverter", "pv")]
)
(await constant(m, "grid", -6000))
(await constant(m, "inverter", 7200))
row = (await m.household_intervals(START, END))["site"][0]
assert row.energy_wh is None
assert row.coverage_seconds == 0
def test_hybrid_rejects_separate_battery_to_avoid_double_counting(setup_measurement):
with pytest.raises(ValidationError):
setup_measurement(
"hybrid_ac",
[dict(key=k, branch=k, role=k) for k in ("grid", "inverter", "battery")],
)
@pytest.mark.asyncio
async def test_missing_ev_does_not_destroy_site(setup_measurement):
m = setup_measurement(inputs=[dict(key=k, branch=k, role=k) for k in ("site", "ev")])
(await constant(m, "site", 800))
rows = (await m.household_intervals(START, END))
assert rows["site"][0].energy_wh == 200
assert rows["household"][0].energy_wh is None
assert rows["household"][0].coverage_seconds == 0
@pytest.mark.asyncio
async def test_intersection_integrates_actual_shape(setup_measurement):
m = setup_measurement(inputs=[dict(key=k, branch=k, role=k) for k in ("site", "ev")])
(await write(m, "site", [(0, 1000, None), (300, 2000, None), (600, None, None)]))
(await write(m, "ev", [(300, 500, None), (900, 500, None)]))
row = (await m.household_intervals(START, END))["household"][0]
assert row.coverage_seconds == 300
assert row.energy_wh is None
assert row.observed_energy_wh == pytest.approx(125)
@pytest.mark.asyncio
async def test_nonoverlapping_coverage_is_missing(setup_measurement):
m = setup_measurement(inputs=[dict(key=k, branch=k, role=k) for k in ("site", "ev")])
(await write(m, "site", [(0, 1000, None), (300, 1000, None)]))
(await write(m, "ev", [(600, 500, None), (900, 500, None)]))
row = (await m.household_intervals(START, END))["household"][0]
assert row.observed_energy_wh is None
assert row.coverage_seconds == 0
@pytest.mark.parametrize(
"inputs",
[
[dict(key="p", branch="x", role="grid"), dict(key="p", branch="y", role="pv")],
[dict(key="p", branch="x", role="grid"), dict(key="q", branch="x", role="pv")],
[dict(key="p", branch="x", role="site"), dict(key="q", branch="y", role="pv")],
[dict(key="unknown", branch="x", role="grid")],
],
)
def test_invalid_balance_configuration(setup_measurement, inputs):
with pytest.raises(ValidationError):
setup_measurement("separate_ac", inputs, {"p": POWER, "q": POWER})
@pytest.mark.asyncio
async def test_quality_reset_even_with_increasing_meter(setup_measurement):
m = setup_measurement(channels={"site": dict(quantity="cumulative_energy", unit="Wh")})
(await write(
m,
"site",
[
(0, 0, {"generation": "old"}),
(450, 1000, {"generation": "new", "reset": True}),
(900, 1100, {"generation": "new", "status": "estimated"}),
],
))
row = (await m.energy_intervals("site", START, END))[0]
assert row.energy_wh is None
assert row.observed_energy_wh == 100
assert set(row.flags) == {"meter_reset", "estimated"}
@pytest.mark.asyncio
async def test_api_upsert_outage_validation_and_readback(setup_measurement):
m = setup_measurement()
app = FastAPI()
app.include_router(router)
with TestClient(app) as client:
payload = [
dict(
date_time=START.isoformat(), key="site", value=800, quality={"status": "estimated"}
),
dict(date_time=END.isoformat(), key="site", value=800),
]
assert client.put("/v1/measurement/samples", json=payload).status_code == 200
params = dict(key="site", start=START.isoformat(), end=END.isoformat())
row = client.get("/v1/measurement/energy", params=params).json()[0]
assert row["energy_wh"] == 200
assert "estimated" in row["flags"]
raw = client.get("/v1/measurement/samples", params=params).json()
assert raw[0]["quality"]["status"] == "estimated"
payload[0]["value"] = None
payload[0]["quality"] = {"status": "unavailable"}
assert client.put("/v1/measurement/samples", json=payload[:1]).status_code == 200
row = client.get("/v1/measurement/energy", params=params).json()[0]
assert row["energy_wh"] is None
assert "unavailable" in row["flags"]
assert len(m.records) == 2
payload[0]["value"] = 0
payload[0]["quality"] = {}
assert client.put("/v1/measurement/samples", json=payload[:1]).status_code == 200
assert client.get("/v1/measurement/energy", params=params).json()[0]["energy_wh"] == 0
assert (
client.get("/v1/measurement/household", params=params).json()["site"][0]["energy_wh"]
== 0
)
# Validate a whole batch before changing any value.
bad = [payload[0] | {"value": 999}, payload[0] | {"key": "unknown"}]
assert client.put("/v1/measurement/samples", json=bad).status_code == 422
assert (await m.energy_intervals("site", START, END))[0].energy_wh == 0
assert (
client.get(
"/v1/measurement/energy", params=params | {"interval_seconds": 0}
).status_code
== 422
)
assert (
client.get(
"/v1/measurement/energy", params=params | {"start": "2026-09-10T00:00:00"}
).status_code
== 422
)
assert (
client.get(
"/v1/measurement/energy",
params=params | {"end": (START + timedelta(days=32)).isoformat()},
).status_code
== 422
)
@pytest.mark.asyncio
@pytest.mark.parametrize("provider", [None, "SQLite", "LMDB"])
async def test_quality_persists_through_storage_restart(
setup_measurement, config_eos, tmp_path, provider
):
m = setup_measurement()
folder, previous_db = config_eos.general.data_folder_path, config_eos.database.provider
try:
config_eos.general.data_folder_path = tmp_path
config_eos.database.provider = provider
(await write(
m,
"site",
[
(0, 800, {"status": "estimated", "generation": "meter1"}),
(450, None, {"status": "unavailable"}),
(900, 800, None),
],
))
before = (await m.energy_intervals("site", START, END))
assert (await m.save())
if provider:
(await get_database().close())
m._db_reset_state()
if not provider:
assert (await m.load())
# DB loads on demand via the bounded query, without a full-history load.
assert (await m.energy_intervals("site", START, END)) == before
assert m.records[0].sample_quality["site"].generation == "meter1"
finally:
if provider:
(await get_database().close())
m._db_reset_state()
config_eos.database.provider = previous_db
config_eos.general.data_folder_path = folder
@pytest.mark.asyncio
async def test_query_passes_bounded_storage_window(setup_measurement, monkeypatch):
m = setup_measurement()
calls = []
original = type(m).db_iterate_records
def record_window(self, start_timestamp=None, end_timestamp=None):
calls.append((start_timestamp, end_timestamp))
return original(self, start_timestamp, end_timestamp)
monkeypatch.setattr(type(m), "db_iterate_records", record_window)
(await m.energy_intervals("site", START, END))
assert calls and all(a is not None and b is not None for a, b in calls)
@pytest.mark.asyncio
async def test_actual_server_routes_and_legacy_value_api(setup_measurement):
setup_measurement()
from akkudoktoreos.server.eos import app
# No lifespan: this test must not start schedulers or write application state.
client = TestClient(app)
try:
for time in (START, END):
response = client.put(
"/v1/measurement/value",
params={
"datetime": time.isoformat(),
"key": "site",
"value": 800,
},
)
assert response.status_code == 200, response.text
response = client.get(
"/v1/measurement/energy",
params={
"key": "site",
"start": START.isoformat(),
"end": END.isoformat(),
},
)
assert response.status_code == 200
assert response.json()[0]["energy_wh"] == 200
assert "sample_quality" not in get_measurement().record_keys
finally:
client.close()
@pytest.mark.asyncio
async def test_quality_merge_keeps_other_channels(setup_measurement):
from akkudoktoreos.measurement.measurement import MeasurementDataRecord
m = setup_measurement(inputs=[dict(key=k, branch=k, role=k) for k in ("site", "ev")])
(await write(m, "site", [(0, 800, {"status": "estimated"})]))
(await write(m, "ev", [(0, None, {"status": "unavailable"})]))
(await m.insert_by_datetime(
MeasurementDataRecord.model_validate(dict(date_time=START, sample_quality={"site": {"status": "measured"}}))
))
assert m.records[0].sample_quality["site"].status == "measured"
assert m.records[0].sample_quality["ev"].status == "unavailable"
@pytest.mark.asyncio
async def test_hour_energy_allocation_and_short_last_interval(setup_measurement):
m = setup_measurement(
channels={
"site": dict(
quantity="interval_energy",
unit="Wh",
interval_seconds=3600,
timestamp_reference="start",
)
}
)
(await write(m, "site", [(0, 1000, None)]))
rows = (await m.household_intervals(START, START + timedelta(seconds=1800)))["site"]
assert [row.energy_wh for row in rows] == [250, 250]
assert all("allocated_energy" in row.methods for row in rows)
row = (await m.energy_intervals("site", START, START + timedelta(seconds=60)))[0]
assert row.coverage_seconds == 60
assert row.energy_wh == pytest.approx(1000 / 60)