From 17e1f5709740e7ef9cc2f2ab9333df022dec3a48 Mon Sep 17 00:00:00 2001
From: AlexKurek <aleksander.kurek@uj.edu.pl>
Date: Tue, 16 Jun 2026 18:41:14 +0200
Subject: [PATCH] Remove unused `np_where` from `rmsimage.py`

---
 bdsf/rmsimage.py | 21 ++++++++-------------
 1 file changed, 8 insertions(+), 13 deletions(-)

diff --git a/bdsf/rmsimage.py b/bdsf/rmsimage.py
index 87c247a..32f95b6 100644
--- a/bdsf/rmsimage.py
+++ b/bdsf/rmsimage.py
@@ -371,9 +371,8 @@ def CheckShape(A):
             if pol == 'I':
                 # Apply mask to mean_map and rms_map by setting masked values to NaN
                 if isinstance(mask, np.ndarray):
-                    pix_masked = np.where(mask == True)
-                    mean[pix_masked] = np.nan
-                    rms[pix_masked] = np.nan
+                    mean[mask] = np.nan
+                    rms[mask] = np.nan
 
                 img.mean_arr = mean
                 img.rms_arr = rms
@@ -400,9 +399,8 @@ def CheckShape(A):
                     else:
                         resdir = img.basedir + '/background/'
                     if not os.path.exists(resdir): os.makedirs(resdir)
-                    zero_pixels = np.where(rms <= 0.0)
                     rms_nonzero = rms.copy()
-                    rms_nonzero[zero_pixels] = np.nan
+                    rms_nonzero[rms <= 0.0] = np.nan
                     func.write_image_to_file(img.use_io, img.imagename + '.norm_I.fits', (image-mean)/rms_nonzero, img, resdir)
                     mylog.info('%s %s' % ('Writing ', resdir+img.imagename+'.norm_I.fits'))
             else:
@@ -423,8 +421,7 @@ def check_rmsmap(self, img, rms):
         bm = (img.beam[0], img.beam[1])
         fw_pix = sqrt(np.prod(bm)/abs(np.prod(cdelt)))
         if img.masked:
-            unmasked = np.where(~img.mask_arr)
-            stdsub = np.std(rms[unmasked])
+            stdsub = np.std(rms[~img.mask_arr])
         else:
             stdsub = np.std(rms)
 
@@ -451,8 +448,7 @@ def check_meanmap(self, img, mean):
         bm = (img.beam[0], img.beam[1])
         fw_pix = sqrt(np.prod(bm)/abs(np.prod(cdelt)))
         if img.masked:
-            unmasked = np.where(~img.mask_arr)
-            stdsub = np.std(mean[unmasked])
+            stdsub = np.std(mean[~img.mask_arr])
         else:
             stdsub = np.std(mean)
         rms_expect = img.clipped_rms/img.rms_box[0]*fw_pix
@@ -763,7 +759,7 @@ def rms_mean_map(self, arr, mask=False, kappa=3, box=None, ncores=None):
             rms_map[co] = cr
 
         # Check if all regions have too few unmasked pixels
-        if mask is not None and np.size(np.where(mean_map != np.inf)) == 0:
+        if mask is not None and not np.any(mean_map != np.inf):
             raise RuntimeError("No unmasked regions from which to determine "\
                          "mean and rms maps")
 
@@ -862,8 +858,7 @@ def correct_borders(map):
 
         # Step 5: fill in boxes with < 5 unmasked pixels (set to values of
         # np.inf)
-        unmasked_boxes = np.where(mean_map != np.inf)
-        if np.size(unmasked_boxes,1) < mapshape[0]*mapshape[1]:
+        if np.count_nonzero(mean_map != np.inf) < mapshape[0]*mapshape[1]:
             mean_map = self.fill_masked_regions(mean_map)
             rms_map = self.fill_masked_regions(rms_map)
 
@@ -902,7 +897,7 @@ def fill_masked_regions(self, themap, magic=np.inf):
                     themap[x, y] = np.nansum(goodcutout)/float(len(goodcutout))
                 delx += 1
                 dely += 1
-        themap[np.where(np.isnan(themap))] = 0.0
+        themap[np.isnan(themap)] = 0.0
         return themap
 
     def pad_array(self, arr, new_shape):