Merge pull request #70 from klarman-cell-observatory/yiming

Add unit test for data aggregation

.github/workflows/ci-test.yml

@@ -39,3 +39,7 @@ jobs:
     - name: IO test
       run: |
         pytest tests/test_io.py
+    - name: Data aggregation test
+      run: |
+        bash tests/run_aggregate_matrix.sh
+        pytest tests/test_aggr.py

pegasusio/data_aggregation.py

@@ -169,7 +169,8 @@ def aggregate_matrices(
         if "Object" in row:
             data = row["Object"].copy()
         else:
-            assert "Location" in row
+            assert "Location" in row, f"Row of sample '{row['Sample']}' must contain a 'Location' column!"
+            assert not row.isnull().values.any(), f"Row of sample '{row['Sample']}' has one or more NaN/NA values!"
 
             input_file = os.path.expanduser(os.path.expandvars(row["Location"].rstrip(os.sep))) # extend all user variables
             file_type, copy_path, copy_type = infer_file_type(input_file) # infer file type
@@ -198,7 +199,7 @@ def aggregate_matrices(
                 genome = def_genome
             modality = row.get("Modality", None)
             data = read_input(input_file, file_type = file_type, genome = genome, modality = modality)
-        
+
         if len(genome_dict) > 0:
             data._update_genome(genome_dict)
 

tests/run_aggregate_matrix.sh

@@ -0,0 +1 @@
+pegasusio aggregate_matrix pegasusio-test-data/aggr_case/count_matrix.csv pegasusio-test-data/aggr_result --attributes Donor --min-genes 100 --mito-prefix MT- --percent-mito 10

tests/test_aggr.py

@@ -0,0 +1,18 @@
+import unittest
+
+import pegasusio as io
+
+class TestAggr(unittest.TestCase):
+
+    def test_aggregate_matrices(self):
+        data = io.read_input("pegasusio-test-data/aggr_result.zarr.zip")
+
+        self.assertEqual(data.shape, (16303, 36601), "Count matrix shape differs!")
+        self.assertTrue('Donor' in data.obs, "Attribute 'Donor' is not included in the resulting count matrix!")
+        self.assertEqual(data.obs['Channel'].cat.categories.size, 3, "Some sample is not included in the result count matrix!")
+        self.assertLessEqual(data.obs['percent_mito'].max(), 10, "Filtration based on '--percent-mito' fails!")
+        self.assertGreaterEqual(data.obs['n_genes'].min(), 100, "Filtration based on '--min-genes' fails!")
+
+
+if __name__ == '__main__':
+    unittest.main()