Merge pull request #1277 from flatironinstitute/dev_no_relative_impor…

…ts_2

Dev no relative imports 2

bin/caiman_gui.py

@@ -12,7 +12,7 @@
 from pyqtgraph.parametertree import Parameter, ParameterTree
 from scipy.sparse import csc_matrix
 
-import caiman as cm
+import caiman
 from caiman.source_extraction.cnmf.cnmf import load_CNMF
 from caiman.source_extraction.cnmf.params import CNMFParams
 
@@ -67,10 +67,10 @@ def make_color_img(img, gain=255, min_max=None, out_type=np.uint8):
                                           directory=d, filter=f + ';;*.mmap')[0]
 
 if fpath[-3:] == 'nwb':
-    mov = cm.load(cnm_obj.mmap_file,
+    mov = caiman.load(cnm_obj.mmap_file,
                   var_name_hdf5='acquisition/TwoPhotonSeries')
 else:
-    mov = cm.load(cnm_obj.mmap_file)
+    mov = caiman.load(cnm_obj.mmap_file)
 
 estimates = cnm_obj.estimates
 params_obj = cnm_obj.params
@@ -106,7 +106,7 @@ def selectionchange(self,i):
     cb.show()
 
 if not hasattr(estimates, 'Cn'):
-    estimates.Cn = cm.local_correlations(mov, swap_dim=False)
+    estimates.Cn = caiman.local_correlations(mov, swap_dim=False)
 #Cn = estimates.Cn
 
 # We rotate our components 90 degrees right because of incompatibility of pyqtgraph and pyplot

caiman/__init__.py

@@ -1,11 +1,10 @@
 #!/usr/bin/env python
 
 import pkg_resources
-from .base.movies import movie, load, load_movie_chain, _load_behavior, play_movie
-from .base.timeseries import concatenate
-from .cluster import start_server, stop_server
-from .mmapping import load_memmap, save_memmap, save_memmap_each, save_memmap_join
-from .summary_images import local_correlations
-#from .source_extraction import cnmf
+from caiman.base.movies import movie, load, load_movie_chain, _load_behavior, play_movie
+from caiman.base.timeseries import concatenate
+from caiman.cluster import start_server, stop_server
+from caiman.mmapping import load_memmap, save_memmap, save_memmap_each, save_memmap_join
+from caiman.summary_images import local_correlations
 
 __version__ = pkg_resources.get_distribution('caiman').version

caiman/base/movies.py

@@ -30,25 +30,18 @@
 import zarr
 from zipfile import ZipFile
 
-import caiman as cm
-
-from . import timeseries
+import caiman.base.timeseries
+import caiman.base.traces
+import caiman.mmapping
+import caiman.summary_images
+import caiman.utils.visualization
 
 try:
     cv2.setNumThreads(0)
 except:
     pass
 
-from . import timeseries as ts
-from .traces import trace
-
-from ..mmapping import load_memmap
-from ..utils import visualization
-from .. import summary_images as si
-from ..motion_correction import apply_shift_online, motion_correct_online
-
-
-class movie(ts.timeseries):
+class movie(caiman.base.timeseries.timeseries):
     """
     Class representing a movie. This class subclasses timeseries,
     that in turn subclasses ndarray
@@ -350,7 +343,7 @@ def extract_shifts(self, max_shift_w: int = 5, max_shift_h: int = 5, template=No
 
         template = template[ms_h:h_i - ms_h, ms_w:w_i - ms_w].astype(np.float32)
 
-        #% run algorithm, press q to stop it
+        # run algorithm, press q to stop it
         shifts = []    # store the amount of shift in each frame
         xcorrs = []
 
@@ -549,11 +542,11 @@ def removeBL(self, windowSize:int=100, quantilMin:int=8, in_place:bool=False, re
         myperc = partial(np.percentile, q=quantilMin, axis=-1)
         res = np.array(list(map(myperc,iter_win))).T
         if returnBL:
-                return cm.movie(cv2.resize(res,pixs.shape[::-1]),fr=self.fr).to3DFromPixelxTime(self.shape)
+                return caiman.movie(cv2.resize(res,pixs.shape[::-1]),fr=self.fr).to3DFromPixelxTime(self.shape)
         if (not in_place):            
             return (pixs-cv2.resize(res,pixs.shape[::-1])).to3DFromPixelxTime(self.shape)
         else:
-            self -= cm.movie(cv2.resize(res,pixs.shape[::-1]),fr=self.fr).to3DFromPixelxTime(self.shape) 
+            self -= caiman.movie(cv2.resize(res,pixs.shape[::-1]),fr=self.fr).to3DFromPixelxTime(self.shape) 
             return self
 
     def to2DPixelxTime(self, order='F'):
@@ -611,7 +604,7 @@ def computeDFF(self, secsWindow: int = 5, quantilMin: int = 8, method: str = 'on
                         **self.__dict__)
         numFramesNew, linePerFrame, pixPerLine = np.shape(mov_out)
 
-        #% compute baseline quickly
+        # compute baseline quickly
         logging.debug("binning data ...")
         sys.stdout.flush()
 
@@ -634,7 +627,7 @@ def computeDFF(self, secsWindow: int = 5, quantilMin: int = 8, method: str = 'on
                                    cval=0.0,
                                    prefilter=False)
 
-        #% compute DF/F
+        # compute DF/F
         if not in_place:
             if method == 'delta_f_over_sqrt_f':
                 mov_out = (mov_out - movBL) / np.sqrt(movBL)
@@ -884,7 +877,7 @@ def local_correlations(self,
         T = self.shape[0]
         Cn = np.zeros(self.shape[1:])
         if T <= 3000:
-            Cn = si.local_correlations(np.array(self),
+            Cn = caiman.summary_images.local_correlations(np.array(self),
                                        eight_neighbours=eight_neighbours,
                                        swap_dim=swap_dim,
                                        order_mean=order_mean)
@@ -894,7 +887,7 @@ def local_correlations(self,
             for jj, mv in enumerate(range(n_chunks - 1)):
                 logging.debug('number of chunks:' + str(jj) + ' frames: ' +
                               str([mv * frames_per_chunk, (mv + 1) * frames_per_chunk]))
-                rho = si.local_correlations(np.array(self[mv * frames_per_chunk:(mv + 1) * frames_per_chunk]),
+                rho = caiman.summary_images.local_correlations(np.array(self[mv * frames_per_chunk:(mv + 1) * frames_per_chunk]),
                                             eight_neighbours=eight_neighbours,
                                             swap_dim=swap_dim,
                                             order_mean=order_mean)
@@ -905,7 +898,7 @@ def local_correlations(self,
 
             logging.debug('number of chunks:' + str(n_chunks - 1) + ' frames: ' +
                           str([(n_chunks - 1) * frames_per_chunk, T]))
-            rho = si.local_correlations(np.array(self[(n_chunks - 1) * frames_per_chunk:]),
+            rho = caiman.summary_images.local_correlations(np.array(self[(n_chunks - 1) * frames_per_chunk:]),
                                         eight_neighbours=eight_neighbours,
                                         swap_dim=swap_dim,
                                         order_mean=order_mean)
@@ -956,7 +949,7 @@ def partition_FOV_KMeans(self,
         fovs = cv2.resize(np.uint8(fovs), (w1, h1), 1. / fx, 1. / fy, interpolation=cv2.INTER_NEAREST)
         return np.uint8(fovs), mcoef, distanceMatrix
 
-    def extract_traces_from_masks(self, masks: np.ndarray) -> trace:
+    def extract_traces_from_masks(self, masks: np.ndarray) -> caiman.base.traces.trace:
         """
         Args:
             masks: array, 3D with each 2D slice bein a mask (integer or fractional)
@@ -975,7 +968,7 @@ def extract_traces_from_masks(self, masks: np.ndarray) -> trace:
 
         pixelsA = np.sum(A, axis=1)
         A = A / pixelsA[:, None]       # obtain average over ROI
-        traces = trace(np.dot(A, np.transpose(Y)).T, **self.__dict__)
+        traces = caiman.base.traces.trace(np.dot(A, np.transpose(Y)).T, **self.__dict__)
         return traces
 
     def resize(self, fx=1, fy=1, fz=1, interpolation=cv2.INTER_AREA):
@@ -1013,7 +1006,7 @@ def resize(self, fx=1, fy=1, fz=1, interpolation=cv2.INTER_AREA):
                 if len(new_m) == 0:
                     new_m = m_tmp
                 else:
-                    new_m = timeseries.concatenate([new_m, m_tmp], axis=0)
+                    new_m = caiman.base.timeseries.concatenate([new_m, m_tmp], axis=0)
 
             return new_m
         else:
@@ -1566,7 +1559,7 @@ def load(file_name: Union[str, list[str]],
 
         elif extension == '.mmap':
             filename = os.path.split(file_name)[-1]
-            Yr, dims, T = load_memmap(
+            Yr, dims, T = caiman.mmapping.load_memmap(
                 os.path.join(                  # type: ignore # same dims typing issue as above
                     os.path.split(file_name)[0], filename))
             images = np.reshape(Yr.T, [T] + list(dims), order='F')
@@ -1666,10 +1659,10 @@ def load_movie_chain(file_list: list[str],
             m = m[:, top:h - bottom, left:w - right, z_top:d - z_bottom]
 
         mov.append(m)
-    return ts.concatenate(mov, axis=0)
+    return caiman.base.timeseries.concatenate(mov, axis=0)
 
 ####
-# This is only used for demo_behavior, and used to be part of cm.load(), activated with the
+# This is only used for demo_behavior, and used to be part of caiman.load(), activated with the
 # 'is_behavior' boolean flag.
 
 def _load_behavior(file_name:str) -> Any:
@@ -1976,7 +1969,7 @@ def from_zip_file_to_movie(zipfile_name: str, start_end:Optional[tuple] = None)
 
                     counter += 1
 
-    return cm.movie(mov[:counter])
+    return caiman.movie(mov[:counter])
 
 
 def from_zipfiles_to_movie_lists(zipfile_name: str, max_frames_per_movie: int = 3000,
@@ -2314,8 +2307,7 @@ def get_file_size(file_name, var_name_hdf5='mov') -> tuple[tuple, Union[int, tup
         else:
             raise Exception('File not found!')
     elif isinstance(file_name, tuple):
-        from ...base.movies import load
-        dims = load(file_name[0], var_name_hdf5=var_name_hdf5).shape
+        dims = caiman.base.movies.load(file_name[0], var_name_hdf5=var_name_hdf5).shape
         T = len(file_name)
 
     elif isinstance(file_name, list):
@@ -2494,7 +2486,7 @@ def animate(i):
         anim = matplotlib.animation.FuncAnimation(fig, animate, frames=frames, interval=1, blit=True)
 
         # call our new function to display the animation
-        return visualization.display_animation(anim, fps=fr)
+        return caiman.utils.visualization.display_animation(anim, fps=fr)
 
     elif backend == 'embed_opencv':
         stopButton = widgets.ToggleButton(

caiman/base/rois.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
 import cv2
 import json
@@ -21,7 +20,7 @@
 from typing import Any, Optional
 import zipfile
 
-from ..motion_correction import tile_and_correct
+from caiman.motion_correction import tile_and_correct
 
 try:
     cv2.setNumThreads(0)
@@ -242,7 +241,7 @@ def nf_match_neurons_in_binary_masks(masks_gt,
     cm_cnmf = [scipy.ndimage.center_of_mass(mm) for mm in masks_comp]
 
     if D is None:
-        #% find distances and matches
+        # find distances and matches
         # find the distance between each masks
         D = distance_masks([A_ben, A_cnmf], [cm_ben, cm_cnmf], min_dist, enclosed_thr=enclosed_thr)
 
@@ -252,7 +251,7 @@ def nf_match_neurons_in_binary_masks(masks_gt,
     matches = matches[0]
     costs = costs[0]
 
-    #%% compute precision and recall
+    # compute precision and recall
     TP = np.sum(np.array(costs) < thresh_cost) * 1.
     FN = np.shape(masks_gt)[0] - TP
     FP = np.shape(masks_comp)[0] - TP
@@ -264,7 +263,7 @@ def nf_match_neurons_in_binary_masks(masks_gt,
     performance['accuracy'] = (TP + TN) / (TP + FP + FN + TN)
     performance['f1_score'] = 2 * TP / (2 * TP + FP + FN)
     logging.debug(performance)
-    #%%
+
     idx_tp = np.where(np.array(costs) < thresh_cost)[0]
     idx_tp_ben = matches[0][idx_tp]    # ground truth
     idx_tp_cnmf = matches[1][idx_tp]   # algorithm - comp
@@ -471,7 +470,7 @@ def register_ROIs(A1,
     matches = matches[0]
     costs = costs[0]
 
-    #%% store indices
+    # store indices
 
     idx_tp = np.where(np.array(costs) < thresh_cost)[0]
     if len(idx_tp) > 0:
@@ -488,7 +487,7 @@ def register_ROIs(A1,
         non_matched1 = list(range(D[0].shape[0]))
         non_matched2 = list(range(D[0].shape[1]))
 
-    #%% compute precision and recall
+    # compute precision and recall
 
     FN = D[0].shape[0] - TP
     FP = D[0].shape[1] - TP
@@ -1001,7 +1000,7 @@ def getfloat():
     fill_color = get32()
     subtype = get16()
     if subtype != 0:
-        raise ValueError('roireader: ROI subtype %s not supported (!= 0)' % subtype)
+        raise ValueError(f'roireader: ROI subtype {subtype} not supported (!= 0)')
     options = get16()
     arrow_style = get8()
     arrow_head_size = get8()

caiman/base/traces.py

@@ -1,25 +1,23 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
-"""
-author: agiovann
-"""
-#%%
+
 import cv2
 import logging
 import numpy as np
 import pylab as pl
 pl.ion()
-from . import timeseries as ts
+
+import caiman.base.timeseries
 
 try:
     cv2.setNumThreads(0)
 except:
     pass
 
-#%%
+################
+# This holds the trace class, which is a specialised Caiman timeseries class.
 
 
-class trace(ts.timeseries):
+class trace(caiman.base.timeseries.timeseries):
     """
     Class representing a trace.
 
@@ -130,7 +128,3 @@ def plot(self, stacked=True, subtract_minimum=False, cmap=pl.cm.jet, **kwargs):
     def extract_epochs(self, trigs=None, tb=1, ta=1):
         raise Exception('Not Implemented. Look at movie resize')
 
-
-if __name__ == "__main__":
-    tracedata = trace(3 + np.random.random((2000, 4)), fr=30, start_time=0)
-    tracedata_dff = tracedata.computeDFF()

caiman/behavior/behavior.py

@@ -1,5 +1,4 @@
 #!/usr/bin/env python
-# -*- coding: utf-8 -*-
 
 """
 Functions related to optical flow
@@ -14,7 +13,7 @@
 from sklearn.decomposition import NMF
 import time
 
-import caiman as cm
+import caiman
 
 try:
     cv2.setNumThreads(0)
@@ -75,14 +74,14 @@ def extract_motor_components_OF(m,
         mask = coo_matrix(np.array(mask).squeeze())
         ms = [get_nonzero_subarray(mask.multiply(fr), mask) for fr in m]
         ms = np.dstack(ms)
-        ms = cm.movie(ms.transpose([2, 0, 1]))
+        ms = caiman.movie(ms.transpose([2, 0, 1]))
 
     else:
         ms = m
     of_or = compute_optical_flow(ms, do_show=False, polar_coord=False)
     of_or = np.concatenate([
-        cm.movie(of_or[0]).resize(resize_fact, resize_fact, 1)[np.newaxis, :, :, :],
-        cm.movie(of_or[1]).resize(resize_fact, resize_fact, 1)[np.newaxis, :, :, :]
+        caiman.movie(of_or[0]).resize(resize_fact, resize_fact, 1)[np.newaxis, :, :, :],
+        caiman.movie(of_or[1]).resize(resize_fact, resize_fact, 1)[np.newaxis, :, :, :]
     ],
                            axis=0)
 
@@ -130,8 +129,8 @@ def extract_magnitude_and_angle_from_OF(spatial_filter_,
             x, y = scipy.signal.medfilt(time_trace, kernel_size=[1, 1]).T
             x = scipy.signal.savgol_filter(x.squeeze(), sav_filter_size, 1)
             y = scipy.signal.savgol_filter(y.squeeze(), sav_filter_size, 1)
-            mag, dirct = to_polar(x - cm.components_evaluation.mode_robust(x),
-                                  y - cm.components_evaluation.mode_robust(y))
+            mag, dirct = to_polar(x - caiman.components_evaluation.mode_robust(x),
+                                  y - caiman.components_evaluation.mode_robust(y))
             dirct = scipy.signal.medfilt(dirct.squeeze(), kernel_size=1).T
 
         # normalize to pixel units
@@ -271,8 +270,8 @@ def compute_optical_flow(m,
 
     return mov_tot
 
+# NMF
 
-#%% NMF
 def extract_components(mov_tot,
                        n_components: int = 6,
                        normalize_std: bool = True,