Merge pull request #936 from flatironinstitute/dev

Merge dev to master to prepare for release

caiman/base/movies.py

@@ -1544,7 +1544,7 @@ def load(file_name: Union[str, List[str]],
                                                                       np.arange(shape[1]))
                                                          ).reshape((len(ts),) + shape[1:])
                         else:
-                            input_arr = tffl.asarray()
+                            input_arr = tffl.asarray(out='memmap')
                             input_arr = input_arr[subindices]
 
                 else:

caiman/source_extraction/cnmf/initialization.py

@@ -1734,7 +1734,7 @@ def onclick(event):
             [ai, ci_raw, ind_success] = extract_ac(data_filtered_box,
                                                    data_raw_box, ind_ctr, patch_dims)
 
-            if (np.sum(ai > 0) < min_pixel) or (not ind_success):
+            if (not ind_success) or (np.sum(ai > 0) < min_pixel):
                 # bad initialization. discard and continue
                 continue
             else:
@@ -1903,6 +1903,8 @@ def extract_ac(data_filtered, data_raw, ind_ctr, patch_dims):
     y_diff = np.concatenate([[-1], np.diff(ci)])
     b = np.median(ci[(y_diff >= 0) * (y_diff < sn)])
     ci -= b
+    if np.isnan(ci.sum()):
+        return None, None, False
 
     # return results
     return ai, ci, True

caiman/source_extraction/cnmf/spatial.py

@@ -184,27 +184,24 @@ def update_spatial_components(Y, C=None, f=None, A_in=None, sn=None, dims=None,
     # we compute the indicator from distance indicator
     ind2_, nr, C, f, b_, A_in = computing_indicator(
         Y, A_in, b_in, C, f, nb, method_exp, dims, min_size, max_size, dist, expandCore, dview)
-
-    # remove components that have a nan
-    ff = np.where(np.isnan(np.sum(C, axis=1)))
-    if np.size(ff) > 0:
-        logging.info("Eliminating nan components: {}".format(ff))
-        ff = ff[0]
-        A_in = csc_column_remove(A_in, list(ff))
-        C = np.delete(C, list(ff), 0)
-
-    # remove empty components    
-    ff = np.where(np.sum(C, axis=1)==0)
+
+    # remove components that are empty or have a nan
+    ff = np.where((np.sum(C, axis=1)==0) + np.isnan(np.sum(C, axis=1)))[0]
     if np.size(ff) > 0:
-        logging.info("Eliminating empty components: {}".format(ff))
-        ff = ff[0]
+        logging.info("Eliminating empty and nan components: {}".format(ff))
         A_in = csc_column_remove(A_in, list(ff))
         C = np.delete(C, list(ff), 0)
-
+        # update indices
+        ind_list = list(range(nr-np.size(ff)))
+        for i in ff:
+            ind_list.insert(i, 0)
+        ind_list = np.array(ind_list, dtype=int)
+        ind2_ = [ind_list[np.setdiff1d(a,ff)] if len(a) else a for a in ind2_]
+
+    nr = np.shape(C)[0]
     if normalize_yyt_one and C is not None:
         C = np.array(C)
-        nr_C = np.shape(C)[0]
-        d_ = scipy.sparse.lil_matrix((nr_C, nr_C))
+        d_ = scipy.sparse.lil_matrix((nr, nr))
         d_.setdiag(np.sqrt(np.sum(C ** 2, 1)))
         A_in = A_in * d_
         C = C/(np.sqrt((C**2).sum(1))[:, np.newaxis] + np.finfo(np.float32).eps)