promote develop → main: per-zone weather (22.4.0)

CHANGELOG.md

@@ -1,3 +1,15 @@
+## [22.4.0-rc.1](https://github.com/sequential-parameter-optimization/spotforecast2-safe/compare/v22.3.0...v22.4.0-rc.1) (2026-06-13)
+
+
+### Features
+
+* **weather:** per-zone weather for the ENTSO-E four-zone pipeline ([260a084](https://github.com/sequential-parameter-optimization/spotforecast2-safe/commit/260a08432177181f8cec78a7d1a679379888a16f))
+
+
+### Documentation
+
+* **tasks:** drop stale n-to-1 task wrapper references from task_multi ([00e9dab](https://github.com/sequential-parameter-optimization/spotforecast2-safe/commit/00e9dabfeba10cc975397aef6aef800630f47cb8))
+
 ## [22.3.0](https://github.com/sequential-parameter-optimization/spotforecast2-safe/compare/v22.2.0...v22.3.0) (2026-06-12)
 
 

MODEL_CARD.md

@@ -7,7 +7,7 @@ This card describes what spotforecast2-safe is, how to use it safely, the condit
 | Field | Value |
 | --- | --- |
 | Name | spotforecast2-safe |
-| Version | 22.3.0 |
+| Version | 22.4.0-rc.1 |
 | Type | Deterministic Python library for time series feature engineering and recursive multi-step forecasting. It performs no training of its own. |
 | Developed by | Thomas Bartz-Beielstein, ORCID [0000-0002-5938-5158](https://orcid.org/0000-0002-5938-5158) |
 | Distributed by | the `sequential-parameter-optimization` GitHub organization |
@@ -18,7 +18,7 @@ This card describes what spotforecast2-safe is, how to use it safely, the condit
 
 The library depends only on numpy, pandas, scikit-learn, lightgbm, numba, pyarrow, requests, feature-engine, holidays, astral, and tqdm. It deliberately excludes plotly, matplotlib, spotoptim, optuna, torch, and tensorflow, so no plotting or automated-tuning code ships in this package.
 
-Two Common Platform Enumeration (CPE) identifiers let vulnerability-tracking and software bill of materials (SBOM) tools recognize the package. The wildcard identifier `cpe:2.3:a:sequential_parameter_optimization:spotforecast2_safe:*:*:*:*:*:*:*:*` matches any release; the current release is `cpe:2.3:a:sequential_parameter_optimization:spotforecast2_safe:22.3.0:*:*:*:*:*:*:*`.
+Two Common Platform Enumeration (CPE) identifiers let vulnerability-tracking and software bill of materials (SBOM) tools recognize the package. The wildcard identifier `cpe:2.3:a:sequential_parameter_optimization:spotforecast2_safe:*:*:*:*:*:*:*:*` matches any release; the current release is `cpe:2.3:a:sequential_parameter_optimization:spotforecast2_safe:22.4.0-rc.1:*:*:*:*:*:*:*`.
 
 The library itself is a low-risk component: it is deterministic, its source is fully inspectable, and it fails safe on invalid input. It is built to support high-risk AI systems in the sense of the EU AI Act, but it is not itself such a system. When it is embedded in a high-risk deployment, the duties that attach to that system fall on the integrator, not on the library.
 
@@ -30,7 +30,7 @@ Responsibilities are divided as follows.
 | Distribution | sequential-parameter-optimization on GitHub | repository issue tracker |
 | Deployment, operation, and audit | the system integrator | defined per deployment |
 
-The current release is 22.3.0, with a stable public interface pinned in `spotforecast2_safe.__init__.__all__`. The full version history, including release dates, is recorded in `CHANGELOG.md` and on the GitHub Releases page; it is maintained automatically by the release pipeline and is not repeated here.
+The current release is 22.4.0-rc.1, with a stable public interface pinned in `spotforecast2_safe.__init__.__all__`. The full version history, including release dates, is recorded in `CHANGELOG.md` and on the GitHub Releases page; it is maintained automatically by the release pipeline and is not repeated here.
 
 ## 2. Intended Use and Scope
 
@@ -216,7 +216,7 @@ Maintainer: Thomas Bartz-Beielstein, ORCID [0000-0002-5938-5158](https://orcid.o
 }
 ```
 
-Or as a formatted reference: Bartz-Beielstein, T. (2026). *spotforecast2-safe: Safety-critical subset of spotforecast2* (Version 22.3.0) [Computer software]. https://github.com/sequential-parameter-optimization/spotforecast2-safe
+Or as a formatted reference: Bartz-Beielstein, T. (2026). *spotforecast2-safe: Safety-critical subset of spotforecast2* (Version 22.4.0-rc.1) [Computer software]. https://github.com/sequential-parameter-optimization/spotforecast2-safe
 
 The technical report (`bart26h/index.qmd`) is the long-form reference for design rationale, compliance mapping, and evaluation protocol.
 

_freeze/docs/reference/configurator.config_entsoe.ConfigEntsoe/execute-results/html.json

@@ -1,8 +1,8 @@
 {
-  "hash": "df358d89e8796124284183fc6efecd0c",
+  "hash": "c2f76276942f6735a97f7843afa6912d",
   "result": {
     "engine": "jupyter",
-    "markdown": "---\ntitle: configurator.config_entsoe.ConfigEntsoe\n---\n\n\n\n```python\nconfigurator.config_entsoe.ConfigEntsoe(\n    country_code='DE',\n    periods=default_periods(),\n    lags_consider=(lambda: list(range(1, 24)))(),\n    train_size=(lambda: pd.Timedelta(days=(3 * 365)))(),\n    end_train_default='2025-12-31 00:00+00:00',\n    delta_val=(lambda: pd.Timedelta(hours=(24 * 7 * 10)))(),\n    predict_size=24,\n    cv_block_size=None,\n    refit_size=7,\n    random_state=314159,\n    n_hyperparameters_trials=20,\n    data_filename='interim/energy_load.csv',\n    targets=None,\n    use_outlier_detection=True,\n    contamination=0.01,\n    imputation_method='weighted',\n    window_size=72,\n    imputation_window_size=None,\n    use_exogenous_features=True,\n    latitude=51.5136,\n    longitude=7.4653,\n    timezone='UTC',\n    state='NW',\n    include_weather_windows=False,\n    include_holiday_features=False,\n    include_holiday_adjacency_features=False,\n    use_population_weighted_weather=False,\n    include_degree_hours=False,\n    include_apparent_temperature=False,\n    degree_hours_base_heating=15.0,\n    degree_hours_base_cooling=22.0,\n    include_ephemeris_features=False,\n    include_day_type_features=False,\n    include_school_holiday_features=False,\n    poly_features_degree=1,\n    max_poly_features=10,\n    poly_mi_n_jobs=-1,\n    poly_mi_sample_size=4000,\n    include_covid_infection_rate=False,\n    include_entsoe_forecast_load=False,\n    include_entsoe_renewable_forecast=False,\n    include_entsoe_net_load=False,\n    include_entsoe_day_ahead_price=False,\n    include_football_match_window=False,\n    include_energy_saving_window=False,\n    index_name='Time (UTC)',\n    bounds=None,\n    verbose=False,\n    cache_home=None,\n    n_trials_optuna=15,\n    n_trials_spotoptim=10,\n    n_initial_spotoptim=5,\n    max_time_spotoptim=None,\n    warm_start_lags=(lambda: list(DEFAULT_WARM_START_LAGS))(),\n    task='lazy',\n    agg_weights=None,\n    forecaster_factory=None,\n    data_loader=None,\n    test_data_loader=None,\n    auto_save_models=True,\n    data_frame_name='default',\n    number_folds=10,\n    on_weather_failure='raise',\n    on_exog_provider_failure='raise',\n    exog_max_gap_hours=0,\n    exog_max_tail_gap_hours=0,\n    exog_provider_window='full',\n    target_qc_range_mw=None,\n    target_qc_step_mw=None,\n    target_qc_window_days=None,\n    target_corruption_policy='abort',\n    target_max_heal_hours=0,\n    target_anchor_zone_hours=168,\n    target_qc_deviation_mw=None,\n    target_qc_deviation_ref=None,\n    target_qc_deviation_slots=2,\n    retrain_max_age=(lambda: pd.Timedelta(days=7))(),\n)\n```\n\nConfiguration for the ENTSO-E forecasting pipeline.\n\nSingle-target counterpart to `ConfigMulti`, used by the ENTSO-E CLI\n(``spotforecast2.tasks.task_entsoe``) and any single-target pipeline routed\nthrough ``spotforecast2.multitask.runner.run(config_cls=ConfigEntsoe)``.\n\n``ConfigEntsoe`` **inherits every field and method of `ConfigMulti`** — so any\nfeature flag added to ``ConfigMulti`` is available here automatically (this\nis what closes the historical feature-flag parity gap structurally, rather\nthan via a hand-maintained mirror). It differs from ``ConfigMulti`` in\nexactly two ways:\n\n- ``index_name`` defaults to ``\"Time (UTC)\"`` (the ENTSO-E CSV time column)\n  instead of ``\"DateTime\"``.\n- it adds ``retrain_max_age`` — the maximum age of a previously trained model\n  before retraining is required (consumed by\n  `spotforecast2_safe.manager.trainer.should_retrain`).\n\nSee `ConfigMulti` for the full field reference (training/validation windows,\nfeature toggles, exogenous-provider flags, target-corruption knobs, …).\n\n## Parameters {.doc-section .doc-section-parameters}\n\n| Name            | Type                                           | Description                                                                                      | Default                            |\n|-----------------|------------------------------------------------|--------------------------------------------------------------------------------------------------|------------------------------------|\n| index_name      | [str](`str`)                                   | Datetime column name used when resetting the index. Defaults to ``\"Time (UTC)\"``.                | `'Time (UTC)'`                     |\n| retrain_max_age | [pd](`pandas`).[Timedelta](`pandas.Timedelta`) | Maximum age of a trained model before a retrain is forced. Defaults to ``pd.Timedelta(days=7)``. | `(lambda: pd.Timedelta(days=7))()` |\n\n## Examples {.doc-section .doc-section-examples}\n\n\n::: {#fdcc96fe .cell execution_count=1}\n``` {.python .cell-code}\nimport pandas as pd\nfrom spotforecast2_safe.configurator.config_entsoe import ConfigEntsoe\nfrom spotforecast2_safe.configurator.config_multi import ConfigMulti\n\nconfig = ConfigEntsoe(country_code=\"DE\")\n# ENTSO-E-specific defaults:\nprint(\"index_name:\", config.index_name)\nprint(\"retrain_max_age:\", config.retrain_max_age)\nassert config.index_name == \"Time (UTC)\"\nassert config.retrain_max_age == pd.Timedelta(days=7)\n\n# Inherits the full ConfigMulti surface, incl. the opt-in feature flags:\nassert isinstance(config, ConfigMulti)\nconfig = ConfigEntsoe(\n    include_ephemeris_features=True,\n    include_day_type_features=True,\n    include_degree_hours=True,\n)\nprint(\"ephemeris:\", config.include_ephemeris_features)\nprint(\"predict_size:\", config.predict_size)\n```\n\n::: {.cell-output .cell-output-stdout}\n```\nindex_name: Time (UTC)\nretrain_max_age: 7 days 00:00:00\nephemeris: True\npredict_size: 24\n```\n:::\n:::\n\n\n",
+    "markdown": "---\ntitle: configurator.config_entsoe.ConfigEntsoe\n---\n\n\n\n```python\nconfigurator.config_entsoe.ConfigEntsoe(\n    country_code='DE',\n    periods=default_periods(),\n    lags_consider=(lambda: list(range(1, 24)))(),\n    train_size=(lambda: pd.Timedelta(days=(3 * 365)))(),\n    end_train_default='2025-12-31 00:00+00:00',\n    delta_val=(lambda: pd.Timedelta(hours=(24 * 7 * 10)))(),\n    predict_size=24,\n    cv_block_size=None,\n    refit_size=7,\n    random_state=314159,\n    n_hyperparameters_trials=20,\n    data_filename='interim/energy_load.csv',\n    targets=None,\n    use_outlier_detection=True,\n    contamination=0.01,\n    imputation_method='weighted',\n    window_size=72,\n    imputation_window_size=None,\n    use_exogenous_features=True,\n    latitude=51.5136,\n    longitude=7.4653,\n    timezone='UTC',\n    state='NW',\n    include_weather_windows=False,\n    include_holiday_features=False,\n    include_holiday_adjacency_features=False,\n    use_population_weighted_weather=False,\n    per_zone_weather=False,\n    zone_weather_locations=None,\n    include_degree_hours=False,\n    include_apparent_temperature=False,\n    degree_hours_base_heating=15.0,\n    degree_hours_base_cooling=22.0,\n    include_ephemeris_features=False,\n    include_day_type_features=False,\n    include_school_holiday_features=False,\n    poly_features_degree=1,\n    max_poly_features=10,\n    poly_mi_n_jobs=-1,\n    poly_mi_sample_size=4000,\n    include_covid_infection_rate=False,\n    include_entsoe_forecast_load=False,\n    include_entsoe_renewable_forecast=False,\n    include_entsoe_net_load=False,\n    include_entsoe_day_ahead_price=False,\n    include_football_match_window=False,\n    include_energy_saving_window=False,\n    index_name='Time (UTC)',\n    bounds=None,\n    verbose=False,\n    cache_home=None,\n    n_trials_optuna=15,\n    n_trials_spotoptim=10,\n    n_initial_spotoptim=5,\n    max_time_spotoptim=None,\n    warm_start_lags=(lambda: list(DEFAULT_WARM_START_LAGS))(),\n    task='lazy',\n    agg_weights=None,\n    forecaster_factory=None,\n    data_loader=None,\n    test_data_loader=None,\n    auto_save_models=True,\n    data_frame_name='default',\n    number_folds=10,\n    on_weather_failure='raise',\n    on_exog_provider_failure='raise',\n    exog_max_gap_hours=0,\n    exog_max_tail_gap_hours=0,\n    exog_provider_window='full',\n    target_qc_range_mw=None,\n    target_qc_step_mw=None,\n    target_qc_window_days=None,\n    target_corruption_policy='abort',\n    target_max_heal_hours=0,\n    target_anchor_zone_hours=168,\n    target_qc_deviation_mw=None,\n    target_qc_deviation_ref=None,\n    target_qc_deviation_slots=2,\n    retrain_max_age=(lambda: pd.Timedelta(days=7))(),\n)\n```\n\nConfiguration for the ENTSO-E forecasting pipeline.\n\nSingle-target counterpart to `ConfigMulti`, used by the ENTSO-E CLI\n(``spotforecast2.tasks.task_entsoe``) and any single-target pipeline routed\nthrough ``spotforecast2.multitask.runner.run(config_cls=ConfigEntsoe)``.\n\n``ConfigEntsoe`` **inherits every field and method of `ConfigMulti`** — so any\nfeature flag added to ``ConfigMulti`` is available here automatically (this\nis what closes the historical feature-flag parity gap structurally, rather\nthan via a hand-maintained mirror). It differs from ``ConfigMulti`` in\nexactly two ways:\n\n- ``index_name`` defaults to ``\"Time (UTC)\"`` (the ENTSO-E CSV time column)\n  instead of ``\"DateTime\"``.\n- it adds ``retrain_max_age`` — the maximum age of a previously trained model\n  before retraining is required (consumed by\n  `spotforecast2_safe.manager.trainer.should_retrain`).\n\nSee `ConfigMulti` for the full field reference (training/validation windows,\nfeature toggles, exogenous-provider flags, target-corruption knobs, …).\n\n## Parameters {.doc-section .doc-section-parameters}\n\n| Name            | Type                                           | Description                                                                                      | Default                            |\n|-----------------|------------------------------------------------|--------------------------------------------------------------------------------------------------|------------------------------------|\n| index_name      | [str](`str`)                                   | Datetime column name used when resetting the index. Defaults to ``\"Time (UTC)\"``.                | `'Time (UTC)'`                     |\n| retrain_max_age | [pd](`pandas`).[Timedelta](`pandas.Timedelta`) | Maximum age of a trained model before a retrain is forced. Defaults to ``pd.Timedelta(days=7)``. | `(lambda: pd.Timedelta(days=7))()` |\n\n## Examples {.doc-section .doc-section-examples}\n\n\n::: {#2f34fa00 .cell execution_count=1}\n``` {.python .cell-code}\nimport pandas as pd\nfrom spotforecast2_safe.configurator.config_entsoe import ConfigEntsoe\nfrom spotforecast2_safe.configurator.config_multi import ConfigMulti\n\nconfig = ConfigEntsoe(country_code=\"DE\")\n# ENTSO-E-specific defaults:\nprint(\"index_name:\", config.index_name)\nprint(\"retrain_max_age:\", config.retrain_max_age)\nassert config.index_name == \"Time (UTC)\"\nassert config.retrain_max_age == pd.Timedelta(days=7)\n\n# Inherits the full ConfigMulti surface, incl. the opt-in feature flags:\nassert isinstance(config, ConfigMulti)\nconfig = ConfigEntsoe(\n    include_ephemeris_features=True,\n    include_day_type_features=True,\n    include_degree_hours=True,\n)\nprint(\"ephemeris:\", config.include_ephemeris_features)\nprint(\"predict_size:\", config.predict_size)\n```\n\n::: {.cell-output .cell-output-stdout}\n```\nindex_name: Time (UTC)\nretrain_max_age: 7 days 00:00:00\nephemeris: True\npredict_size: 24\n```\n:::\n:::\n\n\n",
     "supporting": [
       "configurator.config_entsoe.ConfigEntsoe_files/figure-html"
     ],