Merge pull request #27 from klarman-cell-observatory/boli

Convert added attributes into categorical and recognize nan for Refer…

pegasusio/data_aggregation.py

@@ -1,4 +1,5 @@
 import os
+import numpy as np
 import pandas as pd
 from subprocess import check_call
 from typing import List, Tuple, Dict, Set
@@ -43,7 +44,7 @@ def _parse_restriction_string(rstr: str) -> Tuple[str, bool, Set[str]]:
 
 def _parse_genome_string(genome_str: str) -> Tuple[str, Dict[str, str]]:
     genome = genome_dict = None
-    if genome_str is not None:        
+    if (genome_str is not None) and (not np.isnan(genome_str)):
         if genome_str.find(":") < 0:
             genome = genome_str
         else:
@@ -200,6 +201,8 @@ def aggregate_matrices(
 
     # Merge data
     aggregated_data = aggrData.aggregate()
+    if len(attributes) > 0:
+        aggregated_data._convert_attributes_to_categorical(attributes)
     logger.info(f"Aggregated {tot} files.")
 
     # Delete temporary file

pegasusio/multimodal_data.py

@@ -510,3 +510,8 @@ def _propogate_genome(self) -> None:
                 unidata = self.data.pop(key)
                 unidata.uns["genome"] = genome
                 self.data[unidata.get_uid()] = unidata
+
+
+    def _convert_attributes_to_categorical(self, attributes: Set[str]) -> None:
+        for unidata in self.data.values():
+            unidata._convert_attributes_to_categorical(attributes)

pegasusio/unimodal_data.py

@@ -3,6 +3,7 @@
 from scipy.sparse import csr_matrix
 from collections.abc import MutableMapping
 from copy import deepcopy
+from natsort import natsorted
 from typing import List, Dict, Union, Set, Tuple
 
 import logging
@@ -527,3 +528,9 @@ def _update_barcode_metadata_info(
             self.barcode_metadata[attr] = np.repeat(row[attr], nsample)
 
         self.metadata["_sample"] = sample_name # record sample_name for merging
+
+
+    def _convert_attributes_to_categorical(self, attributes: Set[str]) -> None:
+        for attr in attributes:
+            values = self.barcode_metadata[attr].values
+            self.barcode_metadata[attr] = pd.Categorical(values, categories=natsorted(np.unique(values)))