Fake data Issue #354 (#360) (#369)

* Fake data Issue #354 (#360)

* Fake data Issue #354

* Fake data Issue #354

* Fake data Issue #354

* Changed relative imports and removed pre-commit from requirements.txt

* Using more flexible python3 version to attempt to fix pre-commit.ci build issue

* Written new Unit Test for fake forecast with specified GSP ID

* Written unit tests for all endpoints/routes and modified gsp.py to support new tests

* Removed duplicated Test Cases

* 1st test case for fake environment

* Fixed accidental activation of fake environment in gsp.py module

* Written and tested remaining test cases for is_fake

* Moved isintance check to prior to is_fake condition

* Added 2 Tests to test_national.py and cleaned up some logic in test_gsp test cases

* Modified gsp.py and national.py modules and accompanying test cases to address feedback

* Fixed incorrect for loop iteration through list (should be singular ForecastSQL object) to forecasts object in the test_national.py test cases

* Modified test cases to use NationalForecastValue, ForecastValue, and the ManyForecasts as the return objects

* Modified test cases to uss pytest.fixture() to yield values from db_session

* Possible fix for test_read_latest_all_gsp_normalized() and test_read_latest_all_gsp()

* 1st experiment for test_read_truth_national_gsp() and test_read_forecast_values_gsp()

* 1st experiment with make_fake_gsp_yields()

* 2nd experiment with make_fake_gsp_yields() - modified test_gsp routes

* 3rd experiment with make_fake_gsp_yields() - modified List Comprehension

* 4th experiment with make_fake_gsp_yields() - hard coded _gsp_id_

* Removed yield and fixture

* Experiment: Create a separate tests/fake/test_gsp_fake.py test case module

* split national tests

* role back changes in tests

* fix national test

* rename

* is_fake to is_fake()

* add to readme, don't use any caching in tests

* role back,

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* tests run locally

---------

Co-authored-by: Vikram Pande <[email protected]>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

.pre-commit-config.yaml

@@ -1,5 +1,5 @@
 default_language_version:
-  python: python3.9
+  python: python3
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks

README.md

@@ -88,6 +88,7 @@ Deployment of this service is now done through terraform cloud.
 - `FORECAST_ERROR_HOURS` - using route `/v0/system/GBstatus/check_last_forecast_run` we can check if a forecast has
         been made in the last `FORECAST_ERROR_HOURS` hours
 - `ADJUST_MW_LIMIT` - the maximum the api is allowed to adjust the national forecast by
+- `FAKE` - This allows fake data to be used, rather than connecting to a database
 
 ## Routes to SQL tables
 

src/cache.py

@@ -37,6 +37,8 @@ def remove_old_cache(
             logger.debug(f"Removing {key} from cache, ({value})")
             keys_to_remove.append(key)
 
+    logger.debug(f"Removing {len(keys_to_remove)} keys from cache")
+
     for key in keys_to_remove:
         try:
             last_updated.pop(key)
@@ -129,7 +131,7 @@ def wrapper(*args, **kwargs):  # noqa
                 return response[route_variables]
 
         # use cache
-        logger.debug("Using cache route")
+        logger.debug(f"Using cache route, cache made at {last_updated[route_variables]}")
         return response[route_variables]
 
     return wrapper

src/gsp.py

@@ -4,9 +4,11 @@
 from typing import List, Optional, Union
 
 import structlog
+from dotenv import load_dotenv
 from fastapi import APIRouter, Depends, Request, Security, status
 from fastapi.responses import Response
 from fastapi_auth0 import Auth0User
+from nowcasting_datamodel.fake import make_fake_forecast, make_fake_forecasts, make_fake_gsp_yields
 from nowcasting_datamodel.models import Forecast, ForecastValue, ManyForecasts
 from sqlalchemy.orm.session import Session
 
@@ -32,6 +34,7 @@
 
 logger = structlog.stdlib.get_logger()
 adjust_limit = float(os.getenv("ADJUST_MW_LIMIT", 0.0))
+load_dotenv()
 
 
 router = APIRouter(
@@ -40,7 +43,12 @@
 NationalYield = GSPYield
 
 
-# corresponds to route /v0/solar/GB/gsp/forecast/all
+def is_fake():
+    """Start FAKE environment"""
+    return int(os.environ.get("FAKE", 0))
+
+
+# corresponds to route /v0/solar/GB/gsp/forecast/all/
 @router.get(
     "/forecast/all/",
     response_model=Union[ManyForecasts, List[OneDatetimeManyForecastValues]],
@@ -80,11 +88,17 @@ def get_all_available_forecasts(
     - **end_datetime_utc**: optional end datetime for the query. e.g '2023-08-12 14:00:00+00:00'
     """
 
-    logger.info(f"Get forecasts for all gsps. The option is {historic=} for user {user}")
-
-    if gsp_ids is not None:
+    if isinstance(gsp_ids, str):
         gsp_ids = [int(gsp_id) for gsp_id in gsp_ids.split(",")]
 
+    if is_fake():
+        if gsp_ids is None:
+            gsp_ids = [int(gsp_id) for gsp_id in range(1, GSP_TOTAL)]
+
+        make_fake_forecasts(gsp_ids=gsp_ids, session=session)
+
+    logger.info(f"Get forecasts for all gsps. The option is {historic=} for user {user}")
+
     start_datetime_utc = format_datetime(start_datetime_utc)
     end_datetime_utc = format_datetime(end_datetime_utc)
     creation_limit_utc = format_datetime(creation_limit_utc)
@@ -137,6 +151,10 @@ def get_forecasts_for_a_specific_gsp_old_route(
     user: Auth0User = Security(get_user()),
 ) -> Union[Forecast, List[ForecastValue]]:
     """Redirects old API route to new route /v0/solar/GB/gsp/{gsp_id}/forecast"""
+
+    if is_fake():
+        make_fake_forecast(gsp_id=gsp_id, session=session)
+
     return get_forecasts_for_a_specific_gsp(
         request=request,
         gsp_id=gsp_id,
@@ -185,6 +203,8 @@ def get_forecasts_for_a_specific_gsp(
     - **creation_utc_limit**: optional, only return forecasts made before this datetime.
     returns the latest forecast made 60 minutes before the target time)
     """
+    if is_fake():
+        make_fake_forecast(gsp_id=gsp_id, session=session)
 
     logger.info(f"Get forecasts for gsp id {gsp_id} forecast of forecast with only values.")
     logger.info(f"This is for user {user}")
@@ -251,11 +271,18 @@ def get_truths_for_all_gsps(
     - **start_datetime_utc**: optional start datetime for the query.
     - **end_datetime_utc**: optional end datetime for the query.
     """
-    logger.info(f"Get PV Live estimates values for all gsp id and regime {regime} for user {user}")
 
-    if gsp_ids is not None:
+    if isinstance(gsp_ids, str):
         gsp_ids = [int(gsp_id) for gsp_id in gsp_ids.split(",")]
 
+    if is_fake():
+        if gsp_ids is None:
+            gsp_ids = [int(gsp_id) for gsp_id in range(1, GSP_TOTAL)]
+
+        make_fake_gsp_yields(gsp_ids=gsp_ids, session=session)
+
+    logger.info(f"Get PV Live estimates values for all gsp id and regime {regime} for user {user}")
+
     start_datetime_utc = format_datetime(start_datetime_utc)
     end_datetime_utc = format_datetime(end_datetime_utc)
 
@@ -286,6 +313,10 @@ def get_truths_for_a_specific_gsp_old_route(
     user: Auth0User = Security(get_user()),
 ) -> List[GSPYield]:
     """Redirects old API route to new route /v0/solar/GB/gsp/{gsp_id}/pvlive"""
+
+    if is_fake():
+        make_fake_gsp_yields(gsp_ids=[gsp_id], session=session)
+
     return get_truths_for_a_specific_gsp(
         request=request,
         gsp_id=gsp_id,
@@ -331,6 +362,9 @@ def get_truths_for_a_specific_gsp(
     If not set, defaults to N_HISTORY_DAYS env var, which if not set defaults to yesterday.
     """
 
+    if is_fake():
+        make_fake_forecast(gsp_id=gsp_id, session=session)
+
     logger.info(
         f"Get PV Live estimates values for gsp id {gsp_id} " f"and regime {regime} for user {user}"
     )

src/national.py

@@ -10,6 +10,7 @@
 from elexonpy.api_client import ApiClient
 from fastapi import APIRouter, Depends, HTTPException, Query, Request, Security
 from fastapi_auth0 import Auth0User
+from nowcasting_datamodel.fake import make_fake_forecast, make_fake_gsp_yields
 from nowcasting_datamodel.read.read import get_latest_forecast_for_gsps
 from sqlalchemy.orm.session import Session
 
@@ -43,6 +44,11 @@
 elexon_forecast_api = GenerationForecastApi(api_client)
 
 
+def is_fake():
+    """Start FAKE environment"""
+    return int(os.environ.get("FAKE", 0))
+
+
 @router.get(
     "/forecast",
     response_model=Union[NationalForecast, List[NationalForecastValue]],
@@ -88,6 +94,9 @@ def get_national_forecast(
     """
     logger.debug("Get national forecasts")
 
+    if is_fake:
+        make_fake_forecast(gsp_id=0, session=session)
+
     start_datetime_utc = format_datetime(start_datetime_utc)
     end_datetime_utc = format_datetime(end_datetime_utc)
     creation_limit_utc = format_datetime(creation_limit_utc)
@@ -204,6 +213,9 @@ def get_national_pvlive(
     """
     logger.info(f"Get national PV Live estimates values " f"for regime {regime} for  {user}")
 
+    if is_fake():
+        make_fake_gsp_yields(gsp_ids=[0], session=session)
+
     return get_truth_values_for_a_specific_gsp_from_database(
         session=session, gsp_id=0, regime=regime
     )

src/tests/fake/test_gsp_fake.py

@@ -0,0 +1,88 @@
+from nowcasting_datamodel.models import ForecastValue, LocationWithGSPYields, ManyForecasts
+
+from gsp import GSP_TOTAL, is_fake
+
+
+def test_is_fake_specific_gsp(monkeypatch, api_client, gsp_id=1):
+    """### Test FAKE environment specific _gsp_id_ routes are populating
+    with fake data.
+
+    #### Parameters
+    - **gsp_id**: Please set to any non-zero integer that is <= GSP_TOTAL
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+
+    # Specific _gsp_id_ route/endpoint for successful connection
+    response = api_client.get(f"/v0/solar/GB/gsp/{gsp_id}/forecast")
+    assert response.status_code == 200
+
+    forecast_value = [ForecastValue(**f) for f in response.json()]
+    assert forecast_value is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")
+
+
+def test_is_fake_get_truths_for_a_specific_gsp(monkeypatch, api_client, gsp_id=1):
+    """### Test FAKE environment specific _gsp_id_ routes are populating
+    with fake data.
+
+    #### Parameters
+    - **gsp_id**: Please set to any non-zero integer that is <= GSP_TOTAL
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+
+    # Specific _gsp_id_ route/endpoint for successful connection
+    response = api_client.get(f"/v0/solar/GB/gsp/{gsp_id}/pvlive")
+    assert response.status_code == 200
+
+    forecast_value = [ForecastValue(**f) for f in response.json()]
+    assert forecast_value is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")
+
+
+def test_is_fake_all_available_forecasts(monkeypatch, api_client):
+    """Test FAKE environment for all GSPs are populating
+    with fake data.
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+
+    # Connect to DB endpoint
+    response = api_client.get("/v0/solar/GB/gsp/forecast/all/")
+    assert response.status_code == 200
+
+    all_forecasts = ManyForecasts(**response.json())
+    assert all_forecasts is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")
+
+
+def test_is_fake_get_truths_for_all_gsps(
+    monkeypatch, api_client, gsp_ids=list(range(1, GSP_TOTAL))
+):
+    """Test FAKE environment for all GSPs for yesterday and today
+    are populating with fake data.
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+
+    # Connect to DB endpoint
+    gsp_ids_str = ", ".join(map(str, gsp_ids))
+    response = api_client.get(f"/v0/solar/GB/gsp/pvlive/all?gsp_ids={gsp_ids_str}")
+    assert response.status_code == 200
+
+    all_forecasts = [LocationWithGSPYields(**f) for f in response.json()]
+    assert all_forecasts is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")

src/tests/fake/test_national_fake.py

@@ -0,0 +1,47 @@
+""" Test for main app """
+
+from freezegun import freeze_time
+
+from national import is_fake
+from pydantic_models import NationalForecastValue, NationalYield
+
+
+def test_is_fake_national_all_available_forecasts(monkeypatch, api_client):
+    """Test FAKE environment for all GSPs are populating
+    with fake data.
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+    # Connect to DB endpoint
+    response = api_client.get("/v0/solar/GB/national/forecast")
+    assert response.status_code == 200
+
+    national_forecast_values = [NationalForecastValue(**f) for f in response.json()]
+    assert national_forecast_values is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")
+
+
+# The freeze time is needed so the cahce doesnt interact with the test in test_national.py
+# Ideally we would not have this
+@freeze_time("2021-12-01")
+def test_is_fake_national_get_truths_for_all_gsps(monkeypatch, api_client):
+    """Test FAKE environment for all GSPs for yesterday and today
+    are populating with fake data.
+    """
+
+    monkeypatch.setenv("FAKE", "1")
+    assert is_fake() == 1
+    # Connect to DB endpoint
+    response = api_client.get("/v0/solar/GB/national/pvlive/")
+    assert response.status_code == 200
+
+    print(response.json())
+
+    national_forecast_values = [NationalYield(**f) for f in response.json()]
+    assert national_forecast_values is not None
+
+    # Disable is_fake environment
+    monkeypatch.setenv("FAKE", "0")

test-docker-compose.yml

@@ -21,6 +21,8 @@ services:
       - DB_URL_PV=postgresql://postgres:postgres@postgres_db:5432/postgres
       - GIT_PYTHON_REFRESH=quiet
       - LOG_LEVEL=DEBUG
+      - DELETE_CACHE_TIME_SECONDS=0
+      - CACHE_TIME_SECONDS=0
     command: >
       bash -c "pytest --cov=./src
       && coverage report -m