Accept path-like objects for file inputs in PANDA and PUMA

netZooPy/condor/condor.py

@@ -305,13 +305,14 @@ def matrices(self, c,resolution):
             # Computes weighted biadjacency matrix.
             A = np.matrix(np.zeros((p, q)))
             for edge in self.net.iterrows():
-                A[gn[edge[1][1]], rg[edge[1][0]]] = edge[1][2]
+                row = edge[1]
+                A[gn[row.iloc[1]], rg[row.iloc[0]]] = row.iloc[2]
 
             # Computes node degrees for the nodesets.
             ki = A.sum(1)
             dj = A.sum(0)
             # Computes sum of edges and bimodularity matrix.
-            m = float(sum(ki))
+            m = float(ki.sum())
             B = A - resolution*((ki @ dj) / m)
 
             # d = self.index_dict

netZooPy/panda/io.py

@@ -1,18 +1,19 @@
 from __future__ import print_function
 
 import math
+import os
 import numpy as np
 import sys
 import pandas as pd
 
 
 def load_motif(motif_file):
-    if type(motif_file) is str:
+    if isinstance(motif_file, (str, os.PathLike)):
         # If motif_file is a filename
         motif_data = pd.read_csv(motif_file, sep="\t", header=None)
         motif_tfs = sorted(set(motif_data[0]))
         motif_genes = sorted(set(motif_data[1]))
-    elif type(motif_file) is not str:
+    else:
         # If motif_file is an object
         if motif_file is None:
             # Computation without motif
@@ -37,7 +38,7 @@ def load_motif(motif_file):
 
 
 def load_expression(expression_file, with_header = False, start = 1, end = None):
-    if type(expression_file) is str:
+    if isinstance(expression_file, (str, os.PathLike)):
         # If we pass an expression file, check if we have a 'with header' flag and read it
 
             if with_header:
@@ -54,7 +55,7 @@ def load_expression(expression_file, with_header = False, start = 1, end = None)
             expression_genes = expression_data.index.tolist()
             expression_samples = expression_data.columns.astype(str)
 
-    elif type(expression_file) is not str:
+    else:
         # Pass expression as a dataframe 
         if expression_file is not None:
             if not isinstance(expression_file, pd.DataFrame):

netZooPy/panda/panda.py

@@ -357,14 +357,14 @@ def processData(
 
         ### Loading the PPI
         # #TODO: move this to io
-        if type(ppi_file) is str:
+        if isinstance(ppi_file, (str, os.PathLike)):
             with Timer("Loading PPI data ..."):
                 self.ppi_data = pd.read_csv(ppi_file, sep="\t", header=None)
                 self.ppi_tfs = sorted(
                     set(pd.concat([self.ppi_data[0], self.ppi_data[1]]))
                 )
                 print("Number of PPIs:", self.ppi_data.shape[0])
-        elif type(ppi_file) is not str:
+        else:
             if ppi_file is not None:
                 if not isinstance(ppi_file, pd.DataFrame):
                     raise Exception("Please provide a pandas dataframe for PPI data.")

netZooPy/puma/puma.py

@@ -118,7 +118,7 @@ def __init__(
         with Timer("Loading miR data ..."):
             # If the mir_file is a string the mir list is read from file
             # otherwise the input list is used directly
-            if type(mir_file) is str:
+            if isinstance(mir_file, (str, os.PathLike)):
                 with open(mir_file, "r") as f:
                     miR = f.read().splitlines()
             elif isinstance(mir_file,list):

tests/test_cobra.py

@@ -37,7 +37,8 @@ def test_cobra():
     pd.testing.assert_frame_equal(G, G_gt, rtol=1e-10, check_exact=False)
 
     q = psi.shape[0]
+    X_mean = np.mean(X.to_numpy(), axis=0)
     for i in range(q):
-        C = Q.to_numpy().dot(np.mean(X, axis=0)[i] * np.diag(psi.to_numpy()[i, :])).dot(Q.to_numpy().T)
-        C_gt = Q_gt.to_numpy().dot(np.mean(X, axis=0)[i] * np.diag(psi_gt.to_numpy()[i, :])).dot(Q_gt.to_numpy().T)
+        C = Q.to_numpy().dot(X_mean[i] * np.diag(psi.to_numpy()[i, :])).dot(Q.to_numpy().T)
+        C_gt = Q_gt.to_numpy().dot(X_mean[i] * np.diag(psi_gt.to_numpy()[i, :])).dot(Q_gt.to_numpy().T)
         pd.testing.assert_frame_equal(pd.DataFrame(C), pd.DataFrame(C_gt), rtol=1e-10, check_exact=False)