Merge develop into master v0.2

.github/workflows/cd.yml

@@ -3,46 +3,40 @@ name: 'Continuous Delivery'
 on:
   pull_request:
     branches:
-      - develop
+      - release
       - main
-    paths:
-      - 'challenge/**'
-      - 'scripts/**'
-      - '.github/**'
-      - 'pyproject.toml'
-      - 'poetry.lock'
 
 jobs:
-  run_testing:
+  deploy_and_test:
     runs-on: ubuntu-latest
+
     steps:
-    - uses: actions/checkout@v4
+    - name: checkout code
+      uses: actions/checkout@v4
 
     - name: Set up Google Cloud SDK
       uses: google-github-actions/auth@v1
       with:
-        credentials_json: ${{ secrets.GCP_CREDENTIAL }}
-    - name: Check directory
-      run: |
-        ls
+        credentials_json: ${{ secrets.GCP_CREDENTIALS }}
+
     - name: Set up Python
       uses: actions/setup-python@v2
       with:
         python-version: '3.10.8'
-    - name: Python version
-      run: |
-        python --version
+
     - name: Install poetry and virtualenv
       run: |
         pip install poetry
+
     - name: Install dependencies
       run: |
         poetry config virtualenvs.create false
-        poetry lock --no-update
         poetry install
-    - name: Push to prod
+
+    - name: Push to production
       run: |
         bash scripts/deploy.sh
+
     - name: Run stress test
       run: |
         make stress-test

.github/workflows/ci.yml

@@ -4,43 +4,39 @@ on:
   pull_request:
     branches:
       - develop
+      - release
       - main
-    paths:
-      - 'challenge/**'
-      - 'scripts/**'
-      - '.github/**'
-      - 'pyproject.toml'
-      - 'poetry.lock'
 
 jobs:
   run_testing:
     runs-on: ubuntu-latest
+
     steps:
-    - uses: actions/checkout@v4
-    - name: Check directory
-      run: |
-        ls
-    - name: Set up Python
+    - name: checkout code
+      uses: actions/checkout@v4
+
+    - name: set up Python
       uses: actions/setup-python@v2
       with:
         python-version: '3.10.8'
-    - name: Python version
-      run: |
-        python --version
-    - name: Install poetry and virtualenv
+
+    - name: Install poetry
       run: |
         pip install poetry
+
     - name: Install dependencies
       run: |
         poetry config virtualenvs.create false
-        poetry lock --no-update
         poetry install
+
     - name: Get model
       run:
         python challenge/model.py
+
     - name: Run model test
       run: |
         make model-test
+
     - name: Run api test
       run: |
         make api-test

challenge/api.py

@@ -1,19 +1,13 @@
-import sys
+from datetime import datetime, timezone
 
 import fastapi
 import pandas as pd
 from fastapi import HTTPException
-from pydantic import BaseModel
+from pydantic import BaseModel, validator
 
 from challenge.model import DelayModel
 
-
-def print_to_file(whatever: any):
-    with open("file.txt", "a") as sys.stdout:
-        print(whatever)
-
-
-valid_opera_values = [
+VALID_OPERA_VALUES = [
     "american airlines",
     "air canada",
     "air france",
@@ -39,90 +33,107 @@ def print_to_file(whatever: any):
     "lacsa",
 ]
 
-valid_tipo_vuelo_values = [
+VALID_TIPO_VUELO_VALUES = [
     "I",
     "N",
 ]
 
-valid_mes_values = range(1, 13)
-
-
-def valid_tipo_vuelo(tipo_vuelo: str) -> bool:
-    return tipo_vuelo in valid_tipo_vuelo_values
+VALID_MES_VALUES = range(1, 13)
 
 
-def valid_opera(opera: str) -> bool:
-    return opera in valid_opera_values
-
-
-def valid_mes(mes_value: int) -> bool:
-    return mes_value in valid_mes_values
+app = fastapi.FastAPI()
+model = DelayModel()
+model.load_model("models")
 
 
 class Flight(BaseModel):
     OPERA: str
     TIPOVUELO: str
     MES: int
 
-
-class FlightData(BaseModel):
-    flights: list[Flight]
-
-
-app = fastapi.FastAPI()
-model = DelayModel()
-model.load_model("models")
-
-
-def flight_data_to_pandas(flight_data: FlightData) -> pd.DataFrame:
-    flight_data_dict = {"OPERA": [], "TIPOVUELO": [], "MES": []}
-    for elem in flight_data.flights:
-        if not valid_opera(elem.OPERA.lower()):
+    @validator("OPERA")
+    def valid_opera(cls, opera_value: str):
+        if opera_value.lower() not in VALID_OPERA_VALUES:
             raise HTTPException(
                 status_code=400,
                 detail=(
-                    f"Value for tipo vuelo not valid. Recieved {elem.OPERA},"
-                    f" expected one from {[v for v in valid_opera_values]}"
+                    f"Value for tipo vuelo not valid. Recieved {opera_value}, "
+                    f"expected one from {VALID_OPERA_VALUES}"
                 ),
             )
-        if not valid_tipo_vuelo(elem.TIPOVUELO.capitalize()):
+        return opera_value
+
+    @validator("TIPOVUELO")
+    def valid_tipo_vuelo(cls, tipo_vuelo_value: str):
+        if tipo_vuelo_value.capitalize() not in VALID_TIPO_VUELO_VALUES:
             raise HTTPException(
                 status_code=400,
                 detail=(
-                    f"Value for tipo vuelo not valid. Recieved {elem.TIPOVUELO},"
-                    f" expected one from {[v for v in valid_tipo_vuelo_values]}"
+                    f"Value for tipo vuelo not valid. Recieved {tipo_vuelo_value}, "
+                    f"expected one from {VALID_TIPO_VUELO_VALUES}"
                 ),
             )
-        if not valid_mes(elem.MES):
+        return tipo_vuelo_value
+
+    @validator("MES")
+    def valid_mes(cls, mes_value: int):
+        if mes_value not in VALID_MES_VALUES:
             raise HTTPException(
                 status_code=400,
                 detail=(
-                    f"Value for tipo vuelo not valid. Recieved {elem.MES},"
-                    f" expected one from {valid_mes_values}"
+                    f"Value for tipo vuelo not valid. Recieved {mes_value}, "
+                    f"expected one from {VALID_MES_VALUES}"
                 ),
             )
+        return mes_value
+
+
+class FlightData(BaseModel):
+    flights: list[Flight]
+
+
+def flight_data_to_pandas(flight_data: FlightData) -> pd.DataFrame:
+    flight_data_dict = {"OPERA": [], "TIPOVUELO": [], "MES": []}
+    for elem in flight_data.flights:
         flight_data_dict["OPERA"].append(elem.OPERA)
         flight_data_dict["TIPOVUELO"].append(elem.TIPOVUELO)
         flight_data_dict["MES"].append(elem.MES)
 
     return pd.DataFrame(flight_data_dict)
 
 
+@app.get("/", status_code=200)
+async def root() -> dict:
+    return {
+        "message": (
+            "welcome to the api for predicting flight delay. Use the /health "
+            "endpoint to get server status, and the /predict endpoint to get your "
+            "prediction from input data."
+        )
+    }
+
+
 @app.get("/health", status_code=200)
 async def get_health() -> dict:
     return {"status": "OK"}
 
 
 @app.post("/predict", status_code=200)
 async def post_predict(flight_data: FlightData) -> dict:
-    # get data and convert to pandas dataframe
-
-    flight_data_df = flight_data_to_pandas(flight_data)
-    preprocessed_data = model.preprocess(flight_data_df)
-
-    column_order = model._model.feature_names_in_
-    preprocessed_data = preprocessed_data[column_order]
-
-    pred = model.predict(preprocessed_data)
-
-    return {"predict": pred}
+    try:
+        # get data and convert to pandas dataframe
+        flight_data_df = flight_data_to_pandas(flight_data)
+        preprocessed_data = model.preprocess(flight_data_df)
+
+        # sorts column to feed the model
+        pred = model.predict(preprocessed_data)
+
+        return {"predict": pred}
+    except Exception as e:
+        # there may be exceptions we don't want to send to the clients, so log them in
+        # an internal file for debugging. Just as a cheap solution.
+        with open("error_logs.txt", "a") as f:
+            f.write(f"{datetime.now(timezone.utc)}: encounter error {e}")
+        raise HTTPException(
+            status_code=500, detail="Internal server error during prediction."
+        )