smftools 0.1.7__py3-none-any.whl → 0.2.1__py3-none-any.whl

smftools/preprocessing/invert_adata.py

@@ -1,8 +1,6 @@
 ## invert_adata
 
-# Optional inversion of the adata
-
-def invert_adata(adata):
+def invert_adata(adata, uns_flag='adata_positions_inverted', force_redo=False):
     """
     Inverts the AnnData object along the column (variable) axis.
 
@@ -15,7 +13,13 @@ def invert_adata(adata):
     import numpy as np
     import anndata as ad
 
-    print("🔄 Inverting AnnData along the column axis...")
+    # Only run if not already performed
+    already = bool(adata.uns.get(uns_flag, False))
+    if (already and not force_redo):
+        # QC already performed; nothing to do
+        return adata
+
+    print("Inverting AnnData along the column axis...")
 
     # Reverse the order of columns (variables)
     inverted_adata = adata[:, ::-1].copy()
@@ -26,5 +30,8 @@ def invert_adata(adata):
     # Optional: Store original coordinates for reference
     inverted_adata.var["Original_var_names"] = adata.var_names[::-1]
 
-    print("✅ Inversion complete!")
+    # mark as done
+    inverted_adata.uns[uns_flag] = True
+
+    print("Inversion complete!")
     return inverted_adata

smftools/preprocessing/load_sample_sheet.py

@@ -1,4 +1,10 @@
-def load_sample_sheet(adata, sample_sheet_path, mapping_key_column='obs_names', as_category=True):
+def load_sample_sheet(adata, 
+                      sample_sheet_path, 
+                      mapping_key_column='obs_names',
+                      as_category=True,
+                      uns_flag='sample_sheet_loaded',
+                      force_reload=True
+                        ):
     """
     Loads a sample sheet CSV and maps metadata into the AnnData object as categorical columns.
 
@@ -13,7 +19,13 @@ def load_sample_sheet(adata, sample_sheet_path, mapping_key_column='obs_names',
     """
     import pandas as pd
 
-    print('🔹 Loading sample sheet...')
+    # Only run if not already performed
+    already = bool(adata.uns.get(uns_flag, False))
+    if already and not force_reload:
+        # QC already performed; nothing to do
+        return
+
+    print('Loading sample sheet...')
     df = pd.read_csv(sample_sheet_path)
     df[mapping_key_column] = df[mapping_key_column].astype(str)
     
@@ -25,7 +37,7 @@ def load_sample_sheet(adata, sample_sheet_path, mapping_key_column='obs_names',
 
     value_columns = [col for col in df.columns if col != mapping_key_column]
     
-    print(f'🔹 Appending metadata columns: {value_columns}')
+    print(f'Appending metadata columns: {value_columns}')
     df = df.set_index(mapping_key_column)
 
     for col in value_columns:
@@ -34,5 +46,8 @@ def load_sample_sheet(adata, sample_sheet_path, mapping_key_column='obs_names',
             mapped = mapped.astype('category')
         adata.obs[col] = mapped
 
-    print('✅ Sample sheet metadata successfully added as categories.' if as_category else '✅ Metadata added.')
+    # mark as done
+    adata.uns[uns_flag] = True
+
+    print('Sample sheet metadata successfully added as categories.' if as_category else 'Metadata added.')
     return adata