progress

pipeline/runner.py

@@ -5,22 +5,29 @@ from .postprocessors import check_for_insanity, create_df_tasks
 from .generators import GENERATORS
 from .run import Run
 from .constants import GRAY
+import argparse
 import platformdirs
 import seaborn as sns
 import matplotlib as mpl
 from pathlib import Path
-from typings import Optional
+from typing import Optional
 
 CACHE_DIR = platformdirs.user_cache_dir("econtai")
 
 def run(output_dir: Optional[str] = None):
-    if output_dir is None:
-        output_dir = Path(".")
-
     load_dotenv()
     _setup_graph_rendering()
 
+    if output_dir is None:
+        output_dir = Path("dist/")
+    else:
+        output_dir = Path(output_dir).resolve()
+
+    output_dir.mkdir(parents=True, exist_ok=True)
+
+
     current_run = Run(output_dir=output_dir, cache_dir=CACHE_DIR)
+    current_run.cache_dir.mkdir(parents=True, exist_ok=True)
 
     # Fetchers (fetchers.py)
     current_run.onet_conn, current_run.onet_version = fetch_onet_database(current_run)
@@ -54,3 +61,14 @@ def _setup_graph_rendering():
 
 
     sns.set_style("white")
+
+
+def main():
+    parser = argparse.ArgumentParser(description="Run the econtai pipeline.")
+    parser.add_argument("--output-dir", type=str, help="The directory to write output files to.")
+    args = parser.parse_args()
+    run(output_dir=args.output_dir)
+
+
+if __name__ == "__main__":
+    main()