Fix problem with timeseries qc images (memory leaking)

PUMI/pipelines/func/deconfound.py

@@ -44,6 +44,7 @@ def create_montage(vol_1, vol_2, vol_corrected):
 
         path = str(Path(os.getcwd() + '/fieldmap_correction_comparison.png'))
         plt.savefig(path)
+        plt.close(fig)
         return path
 
     vol_1 = pick_volume('vol_1', volume=volume)

PUMI/pipelines/func/timeseries_extractor.py

@@ -23,7 +23,7 @@ def extract_timeseries_nativespace_qc(wf, **kwargs):
     wf.connect('inputspec', 'timeseries', qc_timeseries, 'timeseries')
     wf.connect('inputspec', 'atlas', qc_timeseries, 'atlas')
     wf.connect('inputspec', 'modules', qc_timeseries, 'modules')
-    wf.connect(qc_timeseries, 'plotfile', 'sinker', 'regTimeseriesQC')
+    wf.connect(qc_timeseries, 'plotfile', 'sinker', 'qc_timeseries')
 
 
 @FuncPipeline(inputspec_fields=['atlas', 'labels', 'modules', 'anat', 'inv_linear_reg_mtrx', 'inv_nonlinear_reg_mtrx',

PUMI/utils.py

@@ -191,6 +191,9 @@ def plot_roi(roi_img, bg_img=None, cut_coords=5, output_file=None, display_mode=
     Here a mosaic plot with 5 columns is generated by default, the background of every plot is black and the default
     cmap has been changed.
 
+    ATTENTION: If save_img is set to False, then you must close the plot yourself!
+    Otherwise this could lead to memory leaks!
+
     Returns:
         plot (nilearn.plotting.displays.OrthoSlicer): Plot object.
         output_file (str): Path to the saved plot (if save_img is False, None is returned).
@@ -204,14 +207,23 @@ def plot_roi(roi_img, bg_img=None, cut_coords=5, output_file=None, display_mode=
     import os
 
     if save_img:
+        # User wants to save the image directly and doesn't want to only get the plot object to modify it further
+        # before saving.
         if output_file is None:
-            roi_img_name = roi_img.split('/')[-1].split('.')[0]
+            roi_img_name = roi_img.split('/')[-1].split('.')[0]  # Get filename without extension.
+            # Plot will be saved in current working directory with the filename (without extension) of ROI image.
+            # but with the suffix '_plot.png'
             output_file = os.path.join(os.getcwd(), roi_img_name + '_plot.png')
         else:
+            # If the user has specified an absolute path, then we can directly use it.
+            # If we have only received a relative path, then we append this relative path to the current working
+            # directory.
             if not output_file.startswith('/'):
                 output_file = os.path.join(os.getcwd(), output_file)
     else:
-        output_file = None  # make sure that no file is created when save_img=False
+        output_file = None
+        # If output_file is None, then nilearn won't save the image.
+        # In this case, the user must close the plot themself! Otherwise this could lead to memory leaks!
 
     plot = plotting.plot_roi(roi_img, bg_img=bg_img, cut_coords=cut_coords, output_file=output_file,
                              display_mode=display_mode, figure=figure, axes=axes, title=title, annotate=annotate,
@@ -286,7 +298,9 @@ def create_coregistration_qc(registered_brain, template, output_file=None, level
     else:
         if not output_file.startswith('/'):
             output_file = os.path.join(os.getcwd(), output_file)
+
     plot.savefig(output_file)
+    plot.close()
 
     return output_file