Improve script to convert CARDINAL dataset to MMCL format to configur…

…e how to name tabular data files

didactic/scripts/cardinal2mmcl.py

@@ -62,6 +62,7 @@ def save_images_and_tabular_data(
     label_tag: TabularAttribute,
     output_dir: Path,
     label_as_a_feature: bool = True,
+    tabular_tag: str = "tabular",
     subsets: Dict[str, Sequence[int]] = None,
 ) -> None:
     """Serializes the images to disk using `torch.save`, and saves tabular data to a CSV file.
@@ -72,6 +73,7 @@ def save_images_and_tabular_data(
         label_tag: Tabular variable to use as the label.
         output_dir: Directory to save the data to.
         label_as_a_feature: Whether to keep the label as a feature (LaaF) in the tabular data.
+        tabular_tag: Tag to use to describe the selection of tabular features in the tabular data files.
         subsets: Optional dictionary mapping indices of images/rows in specific subsets. If provided, the subsets will
             be saved to separate 'pt' and CSV files.
     """
@@ -103,12 +105,12 @@ def save_images_and_tabular_data(
     # Group data by subset if necessary
     if subsets:
         imgs_by_subset = {f"{name}_images": images[idxs] for name, idxs in subsets.items()}
-        tab_by_subset = {f"{name}_tabular": tabular_df.iloc[idxs] for name, idxs in subsets.items()}
+        tab_by_subset = {f"{name}_{tabular_tag}": tabular_df.iloc[idxs] for name, idxs in subsets.items()}
         labels_by_subset = {f"{name}_{label_tag}": labels[idxs] for name, idxs in subsets.items()}
     else:
         imgs_by_subset = {"images": images}
-        tab_by_subset = {"tabular": tabular_df}
-        labels_by_subset = {"labels": labels}
+        tab_by_subset = {tabular_tag: tabular_df}
+        labels_by_subset = {label_tag: labels}
 
     # Save the images as a tensor
     for tag, subset_imgs in imgs_by_subset.items():
@@ -139,6 +141,12 @@ def main():
     parser.add_argument(
         "--tabular_attrs", type=TabularAttribute, nargs="*", help="Tabular attributes to collect and save"
     )
+    parser.add_argument(
+        "--tabular_tag",
+        type=str,
+        default="tabular",
+        help="Tag to use to describe the selection of tabular features in the tabular data files",
+    )
     parser.add_argument(
         "--label_tag",
         type=TabularAttribute,
@@ -162,10 +170,11 @@ def main():
     args = parser.parse_args()
     kwargs = vars(args)
 
-    img_size, norm_bounds, tabular_attrs, label_tag, laaf, imp_rand_state, output_dir, subsets = (
+    img_size, norm_bounds, tabular_attrs, tabular_tag, label_tag, laaf, imp_rand_state, output_dir, subsets = (
         kwargs.pop("img_size"),
         kwargs.pop("norm_bounds"),
         kwargs.pop("tabular_attrs"),
+        kwargs.pop("tabular_tag"),
         kwargs.pop("label_tag"),
         kwargs.pop("laaf"),
         kwargs.pop("imputer_random_state"),
@@ -198,7 +207,9 @@ def main():
         }
 
     # Save the data
-    save_images_and_tabular_data(images, tabular_df, label_tag, output_dir, label_as_a_feature=laaf, subsets=subsets)
+    save_images_and_tabular_data(
+        images, tabular_df, label_tag, output_dir, label_as_a_feature=laaf, tabular_tag=tabular_tag, subsets=subsets
+    )
 
 
 if __name__ == "__main__":