Dockerization done

kellywujy · kellywujy · commit 640afe4cd388 · 2023-06-20T18:14:10.000-07:00
diff --git a/docker/article-relevance/docker-compose.yml b/docker/article-relevance/docker-compose.yml
@@ -7,11 +7,13 @@ services:
       - DOI_PATH=data/article-relevance/raw
       - PARQUET_PATH=data/article-relevance/processed/prediction_parquet
       - N_RECENT=10
+      - MIN_DATE=
+      - MAX_DATE=
       - AUTO_MIN_DATE=False
       - AUTO_CHECK_DUP=False
 
       # Arguments for relevance prediction script
       - DOI_FILE_PATH=data/article-relevance/raw/gdd_api_return.json
-      - MODEL_PATH=data/article-relevance/processed/logistic_regression_model.joblib
+      - MODEL_PATH=models/article-relevance/logistic_regression_model.joblib
       - OUTPUT_PATH=data/article-relevance/processed
       - SEND_XDD=False
diff --git a/models/article-relevance/logistic_regression_model.joblib b/models/article-relevance/logistic_regression_model.joblib
diff --git a/src/article_relevance/gdd_api_query.py b/src/article_relevance/gdd_api_query.py
@@ -72,6 +72,16 @@ def get_new_gdd_articles(output_path,
         get_new_gdd_articles(n_recent_articles = 1000)
     """
 
+    # ======== To handle placeholder for arguments in the docker compose, convert empty str to None ===
+    if n_recent_articles == '':
+         n_recent_articles = None
+    if min_date == '':
+         min_date = None
+    if max_date == '':
+         max_date = None
+    if term == '':
+         term = None
+        
     # ======== Tests for input data type ==========
     if (n_recent_articles is None) and (min_date is None and max_date is None):
             raise ValueError("Either n_recent_articles or a date range should be specified.")
diff --git a/src/article_relevance/relevance_prediction_parquet.py b/src/article_relevance/relevance_prediction_parquet.py
@@ -10,7 +10,7 @@
 Usage: relevance_prediction.py --doi_file_path=<doi_path> --model_path=<model_path> --output_path=<output_path> --send_xdd=<send_xdd>
 
 Options:
-    --doi_file_path=<doi_file_path>                   The path to where the list of DOI is.
+    --doi_file_path=<doi_file_path>         The path to where the list of DOI is.
     --model_path=<model_path>               The path to where the model object is stored.
     --output_path=<output_path>             The path to where the output files will be saved.
     --send_xdd=<send_xdd>                   When True, relevant articles will be sent to xDD through API query. Default is False.
