-
Notifications
You must be signed in to change notification settings - Fork 2
/
visTools.py
3719 lines (3043 loc) · 170 KB
/
visTools.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# -*- coding: utf-8 -*-
"""
Created on Sun Dec 5 16:03:14 2021
@author: Daniel
"""
# def crossSectionGIFsFromTract(streamlines,refAnatT1,saveDir):
#DEPRICATED IN FAVOR OF densityGifsOfTract, which utilizes the existing crossSectionGIFsFromNifti
# import nibabel as nib
# #use dipy to create the density mask
# from dipy.tracking import utils
# import numpy as np
# import wmaPyTools.roiTools
# from nilearn.image import crop_img
# #nilearn.image.resample_img ? to resample output
# croppedReference=crop_img(refAnatT1)
# densityMap=utils.density_map(streamlines, croppedReference.affine, croppedReference.shape)
# densityNifti=nib.nifti1.Nifti1Image(densityMap,croppedReference.affine, croppedReference.header)
# #refuses to plot single slice, single image
# #from nilearn.plotting import plot_stat_map
# #outImg=plot_stat_map(stat_map_img=densityNifti,bg_img=refAnatT1, cut_coords= 1,display_mode='x',cmap='viridis')
# #obtain boundary coords in subject space in order to
# #use plane generation function
# convertedBoundCoords=wmaPyTools.roiTools.subjectSpaceMaskBoundaryCoords(croppedReference)
# #crossSectionGIFsFromNifti
# dimsList=['x','y','z']
# #brute force with matplotlib
# import matplotlib.pyplot as plt
# for iDims in list(range(len(croppedReference.shape))):
# #this assumes that get_zooms returns zooms in subject space and not image space orientation
# # which may not be a good assumption if the orientation is weird
# subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],refAnatT1.header.get_zooms()[iDims])
# #get the desired broadcast shape and delete current dim value
# broadcastShape=list(croppedReference.shape)
# del broadcastShape[iDims]
# #iterate across slices
# for iSlices in list(range(len(subjectSpaceSlices))):
# #set the slice list entry to the appropriate singular value
# currentSlice=wmaPyTools.roiTools.makePlanarROI(croppedReference, subjectSpaceSlices[iSlices], dimsList[iDims])
# #set up the figure
# fig,ax = plt.subplots()
# ax.axis('off')
# #kind of overwhelming to do this in one line
# refData=np.rot90(np.reshape(croppedReference.get_data()[currentSlice.get_data().astype(bool)],broadcastShape),3)
# plt.imshow(refData, cmap='gray', interpolation='nearest')
# #kind of overwhelming to do this in one line
# densityData=np.rot90(np.reshape(densityNifti.get_data()[currentSlice.get_data().astype(bool)],broadcastShape),3)
# plt.imshow(np.ma.masked_where(densityData<1,densityData), cmap='viridis', alpha=.5, interpolation='nearest')
# figName='dim_' + str(iDims) +'_'+ str(iSlices).zfill(3)
# plt.savefig(figName,bbox_inches='tight')
# plt.clf()
# import os
# from PIL import Image
# import glob
# for iDims in list(range(len(croppedReference.shape))):
# dimStem='dim_' + str(iDims)
# img, *imgs = [Image.open(f) for f in sorted(glob.glob(dimStem+'*.png'))]
# img.save(os.path.join(saveDir,dimStem+'.gif'), format='GIF', append_images=imgs,
# save_all=True, duration=len(imgs)*2, loop=0)
# #this worked for a time, not sure why it doesn't take lists now
# [os.remove(iImgFiles) for iImgFiles in sorted(glob.glob(dimStem+'*.png'))]
# #os.remove()
def plotMultiGifsFrom4DNifti(fourDNifti,referenceAnatomy,saveDir):
import os
import nibabel as nib
for iSlices in range(fourDNifti.shape[3]):
outPath=os.path.join(saveDir, str(iSlices))
if not os.path.exists(outPath):
os.makedirs(outPath)
#just try it, maybe it will work
densityNifti=nib.nifti1.Nifti1Image(fourDNifti.get_data()[:,:,:,iSlices],referenceAnatomy.affine,referenceAnatomy.header)
crossSectionGIFsFromNifti(densityNifti,referenceAnatomy,outPath)
def dispersionReport(outDict,streamlines,saveDir,refAnatT1,distanceParameter=3):
"""
Parameters
----------
outDict : TYPE
DESCRIPTION.
saveDir : TYPE
DESCRIPTION.
Returns
-------
None.
"""
import matplotlib.pyplot as plt
import os
import numpy as np
import dipy.tracking.utils as ut
from dipy.tracking.vox2track import streamline_mapping
import nibabel as nib
streamlineMapping=streamline_mapping(streamlines, refAnatT1.affine)
### ugly stuff, but needed for later
dimVoxelDistVals=np.arange(0, 1+distanceParameter*2) -distanceParameter
#if your input was non iso-metric, you would use the voxeldims vector variale to adjust
#the size of the relevant dimensions. we're assuming isometric for this kernel
voxelDims=np.ones(3)
#create a default "footprint" for the filter, e.g. this is the array form of the kernel
x, y, z = np.meshgrid(dimVoxelDistVals, dimVoxelDistVals, dimVoxelDistVals,indexing='ij')
#I didn't abs, but apparently that doesn't cause a problem?
mask_r = x*x*voxelDims[0] + y*y*voxelDims[1] + z*z*voxelDims[2] <= distanceParameter*distanceParameter
#this is what we iterate over per coord
def streamsInImgSpaceWindow(coord, streamlineMapping, mask_r):
#listPosition=sorted(list(streamlineMapping.keys())).index(coord)/len(streamlineMapping)
#print(listPosition, end = "\r")
x2, y2, z2 = np.meshgrid(dimVoxelDistVals+coord[0], dimVoxelDistVals+coord[1], dimVoxelDistVals+coord[2],indexing='ij')
coordsToCheck=np.asarray([ x2, y2, z2])[:,mask_r].T
#try a few different methods here
#set base intersection
#find the roi coords which correspond to voxels within the streamline mask
#validCoords=list(set(list(tuple([tuple(e) for e in coordsToCheck]))) & set(list(streamlineMapping.keys())))
#get the streamline indexes for the streamlines that fall within these coords
#streamIndexes=list(itertools.chain(*[streamlineMapping[iCoords] for iCoords in validCoords]))
#for list
# streamIndexes=[]
# for iCoords in coordsToCheck:
# if tuple(iCoords) in list(streamlineMapping.keys()):
# streamIndexes=streamIndexes+streamlineMapping[tuple(iCoords)]
#try except
#COMICALLY FASTER
streamIndexes=[]
for iCoords in coordsToCheck:
try:
streamIndexes=streamIndexes+streamlineMapping[tuple(iCoords)]
except:
pass
streamIndexes=np.unique(streamIndexes)
return streamIndexes
outputFiles=list(outDict.keys())
for iOutFiles in outputFiles:
outPath=os.path.join(saveDir, iOutFiles)
if not os.path.exists(outPath):
os.makedirs(outPath)
currentData=outDict[iOutFiles].get_data()
smallestNotZero=np.unique(currentData)[1]
crossSectionGIFsFromNifti(outDict[iOutFiles],refAnatT1,outPath, blendOption=False)
#lets mask out the background
[unique, counts] = np.unique(currentData, return_counts=True)
backgroundVal=unique[np.where(np.max(counts)==counts)]
fig = plt.hist(np.ravel(currentData[currentData!=backgroundVal]),bins=2000)
#fig = plt.hist(np.ravel(currentData[currentData!=backgroundVal]))
plt.title(iOutFiles)
plt.xlabel("value")
plt.ylabel("Frequency")
plt.savefig(os.path.join(outPath,iOutFiles+".png"),dpi=300)
#plot lowest and highest
densityMap=ut.density_map(streamlines, refAnatT1.affine, refAnatT1.shape)
countWeightedValues=np.nan_to_num(np.divide(currentData,densityMap))
#highest first
#find where that location is
maxLocation=np.where(countWeightedValues==np.max(countWeightedValues))
#find the streams in that window
maxLocationList=[maxLocation[0][0],maxLocation[1][0],maxLocation[2][0]]
maxLocationIndexes= streamsInImgSpaceWindow(maxLocationList, streamlineMapping, mask_r)
#name the plot
tractName=os.path.join(outPath,'_'.join([str(x) for x in maxLocationList]) +'_maxValStreams')
dipyPlotTract(streamlines[maxLocationIndexes],refAnatT1=None, tractName=tractName,endpointColorDensityKernel=7)
#
smallestNotZero=np.unique(countWeightedValues)[1]
minLocation=np.where(countWeightedValues==np.min(countWeightedValues[countWeightedValues>=smallestNotZero]))
#find the streams in that window
minLocationList=[minLocation[0][0],minLocation[1][0],minLocation[2][0]]
minLocationIndexes= streamsInImgSpaceWindow(minLocationList, streamlineMapping, mask_r)
#name the plot
tractName=os.path.join(outPath,'_'.join([str(x) for x in minLocationList]) +'_minValStreams')
dipyPlotTract(streamlines[minLocationIndexes],refAnatT1=None, tractName=tractName,endpointColorDensityKernel=7)
#save the nfitis down
nib.save(outDict[iOutFiles],os.path.join(outPath, iOutFiles+'.nii.gz'))
#find highest and lowest streamline-count-weighted values
# get a streamline index dict of the whole brain tract
def generateAnatOverlayXSections(overlayNifti,refAnatT1,saveDir):
import nibabel as nib
#use dipy to create the density mask
from dipy.tracking import utils
import numpy as np
from glob import glob
import os
from nilearn.image import reorder_img
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
from matplotlib import figure
from nilearn.image import crop_img, resample_img
import matplotlib.pyplot as plt
import dipy.tracking.utils as ut
import wmaPyTools.roiTools
#do some testing here for binary vs count data
#RAS reoreintation
refAnatT1=reorder_img(refAnatT1)
overlayNifti=reorder_img(overlayNifti)
#get the resolution of each nifti
overlayRes=overlayNifti.header.get_zooms()
refT1Res=refAnatT1.header.get_zooms()
#if they aren't the same resolution
if not np.array_equal(overlayRes,refT1Res):
#resample the *T1* not the overlay
#this is because this is just for visualization purposes, not quantative
#and resampling the typically sparse overlay leads to substantial losses
refAnatT1=resample_img(refAnatT1,target_affine=overlayNifti.affine[0:3,0:3])
#get their shapes as well
overlayShape=overlayNifti.shape
refT1Shape=refAnatT1.shape
#obtain boundary coords in subject space in order to
#use plane generation function
convertedBoundCoords=wmaPyTools.roiTools.subjectSpaceMaskBoundaryCoords(refAnatT1)
dimsList=['x','y','z']
#brute force with matplotlib
for iDims in list(range(len(refT1Shape))):
if refT1Shape[iDims]<=overlayShape[iDims]:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],refT1Res[iDims])
else:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],overlayRes[iDims])
#get the desired broadcast shape and delete current dim value
broadcastShape=list(refAnatT1.shape)
del broadcastShape[iDims]
#iterate across slices
for iSlices in list(range(len(subjectSpaceSlices))):
#set the slice list entry to the appropriate singular value
#THE SOLUTION WAS SO OBVIOUS. DONT USE A SINGLE SLICE FOR BOTH THE
#REFERNCE AND THE OVERLAY. DUH!
#actually this doesn't matter if we resample
currentRefSlice=wmaPyTools.roiTools.makePlanarROI(refAnatT1, subjectSpaceSlices[iSlices], dimsList[iDims])
#could be an issue if overlay nifti is signifigantly smaller
currentOverlaySlice=wmaPyTools.roiTools.makePlanarROI(overlayNifti, subjectSpaceSlices[iSlices], dimsList[iDims])
#set up the figure
fig,ax = plt.subplots()
ax.axis('off')
#kind of overwhelming to do this in one line
refData=np.rot90(np.reshape(refAnatT1.get_data()[currentRefSlice.get_data().astype(bool)],broadcastShape),1)
plt.imshow(refData, cmap='gray', interpolation='antialiased')
#kind of overwhelming to do this in one line
overlayData=np.rot90(np.reshape(overlayNifti.get_data()[currentOverlaySlice.get_data().astype(bool)],broadcastShape),1)
#lets mask out the background
[unique, counts] = np.unique(overlayData, return_counts=True)
backgroundVal=unique[np.where(np.max(counts)==counts)[0]]
plt.imshow(np.ma.masked_where(overlayData==backgroundVal,overlayData), cmap='jet', alpha=.75, interpolation='antialiased',vmin=0,vmax=np.nanmax(overlayNifti.get_data()))
curFig=plt.gcf()
cbaxes = inset_axes(curFig.gca(), width="5%", height="80%", loc=5)
plt.colorbar(cax=cbaxes, ticks=[0.,np.nanmax(overlayNifti.get_data())], orientation='vertical')
#put some text about the current dimension and slice
yLims=curFig.gca().get_ylim()
#xLims=curFig.gca().get_xlim()
#this changes the resolution of the figure itself, and breaks everything
#fix later
# curFig.gca().text(0, yLims[0]-1, dimsList[iDims] + ' = ' + str(subjectSpaceSlices[iSlices]),
# bbox={'facecolor': 'white', 'alpha': 1, 'pad': 10})
curFig.gca().yaxis.set_ticks_position('left')
curFig.gca().tick_params( colors='white')
# we use *2 in order to afford room for the subsequent blended images
figName='dim_' + str(iDims) +'_'+ str(iSlices).zfill(3)+'.png'
plt.savefig(os.path.join(saveDir,figName),bbox_inches='tight',pad_inches=0.0)
plt.close()
def multiTileOverlay_wrap(overlayNifti,refAnatT1,saveDir,figName,noEmpties=True,postClean=True):
import nibabel as nib
#use dipy to create the density mask
from dipy.tracking import utils
import numpy as np
from glob import glob
import os
from nilearn.image import reorder_img
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
from matplotlib import figure
from nilearn.image import crop_img, resample_img
import matplotlib.pyplot as plt
import dipy.tracking.utils as ut
import wmaPyTools.roiTools
from dipy.segment.mask import bounding_box
refAnatT1=reorder_img(refAnatT1)
overlayNifti=reorder_img(overlayNifti)
generateAnatOverlayXSections(overlayNifti,refAnatT1,saveDir)
imgSpaceMins, imgSpaceMaxs=bounding_box(overlayNifti.get_data())
overlayShape=overlayNifti.shape
refT1Shape=refAnatT1.shape
import os
from PIL import Image
from glob import glob
#create blened images to smooth transitions between slices
for iDims in list(range(len(refAnatT1.shape))):
#presumes continuous slice numbering
if noEmpties:
if refT1Shape[iDims]<=overlayShape[iDims]:
imgSpaceSlices=np.arange(imgSpaceMins[iDims],imgSpaceMaxs[iDims])
else:
#I don't think this can happen
imgSpaceSlices=np.arange(0,refT1Shape[iDims])
else:
imgSpaceSlices=np.arange(0,refT1Shape[iDims])
dimStem='dim_' + str(iDims)
#open the images
img, *imgs = [Image.open(os.path.join(saveDir,dimStem+'_'+str(f).zfill(3)+'.png')) for f in imgSpaceSlices]
#find the number you'll be working with
numImagesToTile=len(imgSpaceSlices)
#find the lenght of the aggregated sides
squareSide=np.ceil(np.sqrt(numImagesToTile)).astype(int)
#get the dimensions of the images in this dimension
imgPix=img.size
#create a blank output array
tileOut=np.zeros([imgPix[0]*squareSide,imgPix[1]*squareSide,np.asarray(img).shape[2]-1]).astype(np.uint8)
#create the iterated bounds
colBounds=np.arange(0,tileOut.shape[0],imgPix[0])
rowBounds=np.arange(0,tileOut.shape[1],imgPix[1])
#get the appropriate number of these to iterate across all imgs
y, x=np.meshgrid(colBounds,rowBounds,indexing='ij')
#turn into 1d vectors
yBoundVec=np.ravel(y)
xBoundVec=np.ravel(x)
for iImgs in range(len(imgs)):
#tileOut[yBoundVec[iImgs]:yBoundVec[iImgs]+imgPix[1],xBoundVec[iImgs]:xBoundVec[iImgs]+imgPix[0],:]=np.transpose(np.asarray(imgs[iImgs]),[1,0,2]).astype(np.uint8)[:,:,0:3]
tileOut[yBoundVec[iImgs]:yBoundVec[iImgs]+imgPix[0],xBoundVec[iImgs]:xBoundVec[iImgs]+imgPix[1],:]=np.transpose(np.asarray(imgs[iImgs]),[1,0,2]).astype(np.uint8)[:,:,0:3]
im=Image.fromarray(np.fliplr(np.rot90(tileOut,3)))
im.save(os.path.join(saveDir,figName+'_'+dimStem+'.png'), format='png')
plt.close('all')
#conditionally remove the generated files
if postClean:
[os.remove(ipaths) for ipaths in sorted(glob(os.path.join(saveDir,dimStem+'_*.png')))]
def multiTileDensity(streamlines,refAnatT1,saveDir,tractName,densityThreshold=0,noEmpties=True):
"""
Parameters
----------
overlayNifti : TYPE
DESCRIPTION.
refAnatT1 : TYPE
DESCRIPTION.
saveDir : TYPE
DESCRIPTION.
noEmpties : TYPE, optional
DESCRIPTION. The default is True.
Returns
-------
None.
"""
import nibabel as nib
#use dipy to create the density mask
from dipy.tracking import utils
import numpy as np
from glob import glob
import os
from nilearn.image import reorder_img
#has to be done before matplotlib import
#check to see if in docker container
import wmaPyTools.genUtils
if wmaPyTools.genUtils.is_docker():
import matplotlib as mpl
mpl.use('Agg')
# print('Docker execution detected\nUsing xvfbwrapper for virtual display')
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# from xvfbwrapper import Xvfb
# vdisplay = Xvfb()
# vdisplay.start()
#stop code
# if wmaPyTools.genUtils.is_docker():
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# vdisplay.stop()
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
from matplotlib import figure
from nilearn.image import crop_img, resample_img
import matplotlib.pyplot as plt
import dipy.tracking.utils as ut
import wmaPyTools.roiTools
#Crop initial T1
#crop the anatomy back down in case it has gotten overly widened
refAnatT1=crop_img(refAnatT1)
#get tract density nifti
tractDensityNifti=ut.density_map(streamlines, refAnatT1.affine, refAnatT1.shape)
#apply the density threshold
tractDensityNifti[tractDensityNifti<(densityThreshold*np.max(tractDensityNifti))]=0
#regenerate the nifti
densityNifti = nib.nifti1.Nifti1Image(tractDensityNifti, refAnatT1.affine, refAnatT1.header)
#RAS reoreintation
refAnatT1=reorder_img(refAnatT1)
overlayNifti=reorder_img(densityNifti)
#obtain boundary coords in subject space in order to
#use plane generation function
convertedBoundCoords=wmaPyTools.roiTools.subjectSpaceMaskBoundaryCoords(refAnatT1)
#overlayBoundCoords=subjectSpaceMaskBoundaryCoords(overlayNifti)
#find which input has the highest resolution
refAnatT1Resolution=refAnatT1.header.get_zooms()
overlayResolution=overlayNifti.header.get_zooms()
#not necessary with density mask
# #change datatype so that nilearn doesn't mess up
# #it doesn't like bool so we have to do something about that
# #somehow the data can be set to nifti, but can't be extracted as nifti, so...
# overlayNifti=nib.nifti1.Nifti1Image(overlayNifti.get_data(),overlayNifti.affine,overlayNifti.header)
# overlayNifti.set_data_dtype(overlayNifti.get_data().dtype)
#resample to the best resolution
#but also assuming that the refernce anatomy is ultimately the shape that we want
# #this is going to cause huge problems if the user passes in a super cropped overlay
# THIS SHOULDN'T BE NECESSARY DUE TO THE USE OF THE REFT1 FOR DENSITY MAP
# if np.prod(refAnatT1Resolution)>np.prod(overlayResolution):
# print('resampling reference anatomy to overlay')
# refAnatT1=resample_img(refAnatT1,target_affine=overlayNifti.affine[0:3,0:3])
# overlayNifti=resample_img(overlayNifti,target_affine=overlayNifti.affine[0:3,0:3])
# else:
# print('resampling overlay to reference anatomy')
# refAnatT1=resample_img(refAnatT1,target_affine=refAnatT1.affine[0:3,0:3])
# overlayNifti=resample_img(overlayNifti,target_affine=refAnatT1.affine[0:3,0:3])
# #crop the anatomy back down in case it has gotten overly widened
# refAnatT1=crop_img(refAnatT1)
# #now crop the overlay to the dimensions of the reference anatomy
# overlayNifti=resample_img(overlayNifti,target_affine=refAnatT1.affine, target_shape=refAnatT1.shape)
#compute the bounds of the actual density data
streamDensityBoundCoords=wmaPyTools.roiTools.subjectSpaceMaskBoundaryCoords(overlayNifti)
dimsList=['x','y','z']
#brute force with matplotlib
for iDims in list(range(len(refAnatT1.shape))):
#this assumes that get_zooms returns zooms in subject space and not image space orientation
# which may not be a good assumption if the orientation is weird
if noEmpties:
# + and - 1 in odrer to provide a pad
#NO PADDING, PADDING IS BAD
if refAnatT1Resolution[iDims]<=overlayResolution[iDims]:
subjectSpaceSlices=np.arange(streamDensityBoundCoords[0,iDims],streamDensityBoundCoords[1,iDims],refAnatT1Resolution[iDims])
else:
subjectSpaceSlices=np.arange(streamDensityBoundCoords[0,iDims],streamDensityBoundCoords[1,iDims],overlayResolution[iDims])
#pick whichever input has the best resolution in this dimension
else:
if refAnatT1Resolution[iDims]<=overlayResolution[iDims]:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],refAnatT1Resolution[iDims])
else:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],overlayResolution[iDims])
#get the desired broadcast shape and delete current dim value
broadcastShape=list(refAnatT1.shape)
del broadcastShape[iDims]
#iterate across slices
for iSlices in list(range(len(subjectSpaceSlices))):
#set the slice list entry to the appropriate singular value
#THE SOLUTION WAS SO OBVIOUS. DONT USE A SINGLE SLICE FOR BOTH THE
#REFERNCE AND THE OVERLAY. DUH!
#actually this doesn't matter if we resample
currentRefSlice=wmaPyTools.roiTools.makePlanarROI(refAnatT1, subjectSpaceSlices[iSlices], dimsList[iDims])
#could be an issue if overlay nifti is signifigantly smaller
currentOverlaySlice=wmaPyTools.roiTools.makePlanarROI(overlayNifti, subjectSpaceSlices[iSlices], dimsList[iDims])
#set up the figure
fig,ax = plt.subplots()
ax.axis('off')
#kind of overwhelming to do this in one line
refData=np.rot90(np.reshape(refAnatT1.get_data()[currentRefSlice.get_data().astype(bool)],broadcastShape),1)
plt.imshow(refData, cmap='gray', interpolation='gaussian')
#kind of overwhelming to do this in one line
overlayData=np.rot90(np.reshape(overlayNifti.get_data()[currentOverlaySlice.get_data().astype(bool)],broadcastShape),1)
#lets mask out the background
[unique, counts] = np.unique(overlayData, return_counts=True)
backgroundVal=unique[np.where(np.max(counts)==counts)[0]]
plt.imshow(np.ma.masked_where(overlayData==backgroundVal,overlayData), cmap='jet', alpha=.75, interpolation='kaiser',vmin=0,vmax=np.nanmax(overlayNifti.get_data()))
curFig=plt.gcf()
yLims=sorted(curFig.gca().get_ylim())
xLims=sorted(curFig.gca().get_xlim())
#this changes the resolution of the figure itself, and breaks everything
#fix later
#also, it's all wonky, so the min and max as well as axis orienations are flipped
curFig.gca().text((yLims[1])*.02, (yLims[1])*.08, dimsList[iDims] + ' = ' + str(np.round(subjectSpaceSlices[iSlices],1)), color='white', fontsize=15)
# bbox={'facecolor': 'white', 'alpha': 1, 'pad': 10})
#be careful here, modifying the cbaxes modifies the current figure
cbaxes = inset_axes(curFig.gca(), width="5%", height="80%", loc=5)
plt.colorbar(cax=cbaxes, ticks=[0.,np.nanmax(overlayNifti.get_data())], orientation='vertical')
#put some text about the current dimension and slice
curFig.gca().yaxis.set_ticks_position('left')
curFig.gca().tick_params(colors='white')
# we use *2 in order to afford room for the subsequent blended images
figName='dim_' + str(iDims) +'_'+ str(iSlices*2).zfill(3)+'.png'
plt.savefig(figName,bbox_inches='tight',pad_inches=0.0)
#this only clears
#plt.clf()
plt.close()
import os
from PIL import Image
from glob import glob
#create blened images to smooth transitions between slices
for iDims in list(range(len(refAnatT1.shape))):
dimStem='dim_' + str(iDims)
#open the images
img, *imgs = [Image.open(f) for f in sorted(glob(dimStem+'_*.png'))]
#find the number you'll be working with
numImagesToTile=len(sorted(glob(dimStem+'_*.png')))
#find the lenght of the aggregated sides
squareSide=np.ceil(np.sqrt(numImagesToTile)).astype(int)
#get the dimensions of the images in this dimension
imgPix=img.size
#create a blank output array
tileOut=np.zeros([imgPix[0]*squareSide,imgPix[1]*squareSide,np.asarray(img).shape[2]-1]).astype(np.uint8)
#create the iterated bounds
colBounds=np.arange(0,tileOut.shape[0],imgPix[0])
rowBounds=np.arange(0,tileOut.shape[1],imgPix[1])
#get the appropriate number of these to iterate across all imgs
y, x=np.meshgrid(colBounds,rowBounds,indexing='ij')
#turn into 1d vectors
yBoundVec=np.ravel(y)
xBoundVec=np.ravel(x)
for iImgs in range(len(imgs)):
#tileOut[yBoundVec[iImgs]:yBoundVec[iImgs]+imgPix[1],xBoundVec[iImgs]:xBoundVec[iImgs]+imgPix[0],:]=np.transpose(np.asarray(imgs[iImgs]),[1,0,2]).astype(np.uint8)[:,:,0:3]
tileOut[yBoundVec[iImgs]:yBoundVec[iImgs]+imgPix[0],xBoundVec[iImgs]:xBoundVec[iImgs]+imgPix[1],:]=np.transpose(np.asarray(imgs[iImgs]),[1,0,2]).astype(np.uint8)[:,:,0:3]
im=Image.fromarray(np.fliplr(np.rot90(tileOut,3)))
im.save(os.path.join(saveDir,tractName+'_'+dimStem+'.png'), format='png')
plt.close('all')
[os.remove(ipaths) for ipaths in sorted(glob(dimStem+'*.png'))]
#stop code
# if wmaPyTools.genUtils.is_docker():
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# vdisplay.stop()
def crossSectionGIFsFromOverlay(overlayNifti,refAnatT1,saveDir,figName,postClean=True):
import matplotlib.pyplot as plt
generateAnatOverlayXSections(overlayNifti,refAnatT1,saveDir)
import os
from PIL import Image
from glob import glob
for iDims in list(range(len(refAnatT1.shape))):
dimStem='dim_' + str(iDims)
img, *imgs = [Image.open(f) for f in sorted(glob(os.path.join(saveDir,dimStem+'_*.png')))]
img.save(os.path.join(saveDir,figName+'_'+dimStem+'.gif'), format='GIF', append_images=imgs,
save_all=True, duration=1, loop=0)
plt.close('all')
if postClean:
[os.remove(ipaths) for ipaths in sorted(glob(os.path.join(saveDir,dimStem+'_*.png')))]
def crossSectionGIFsFromNifti(overlayNifti,refAnatT1,saveDir, blendOption=False):
import nibabel as nib
#use dipy to create the density mask
from dipy.tracking import utils
import numpy as np
from glob import glob
import os
from nilearn.image import reorder_img
#has to be done before matplotlib import
#check to see if in docker container
import wmaPyTools.genUtils
if wmaPyTools.genUtils.is_docker():
import matplotlib as mpl
mpl.use('Agg')
# print('Docker execution detected\nUsing xvfbwrapper for virtual display')
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# from xvfbwrapper import Xvfb
# vdisplay = Xvfb()
# vdisplay.start()
#stop code
# if wmaPyTools.genUtils.is_docker():
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# vdisplay.stop()
from mpl_toolkits.axes_grid1.inset_locator import inset_axes
from matplotlib import figure
from nilearn.image import crop_img, resample_img
import matplotlib.pyplot as plt
import wmaPyTools.roiTools
#resample crop (doesn't seem to work)
#[refAnatT1,overlayNifti]=dualCropNifti(refAnatT1,overlayNifti)
#ok, but if we assume that the overlay is *always* going to be smaller than the
#reference, we can crop the overlay safely and work adaptively. Why would
#you even have overlay **outside** of the refernce
#nilearn doesn't handle NAN gracefully, so we have to be inelegant
#overlayNifti=nib.nifti1.Nifti1Image(np.nan_to_num(overlayNifti.get_data()), overlayNifti.affine, overlayNifti.header)
#croppedOverlayNifti=crop_img(overlayNifti)
#RAS reoreintation
refAnatT1=reorder_img(refAnatT1)
overlayNifti=reorder_img(overlayNifti)
#refuses to plot single slice, single image
#from nilearn.plotting import plot_stat_map
#outImg=plot_stat_map(stat_map_img=densityNifti,bg_img=refAnatT1, cut_coords= 1,display_mode='x',cmap='viridis')
#obtain boundary coords in subject space in order to
#use plane generation function
convertedBoundCoords=wmaPyTools.roiTools.subjectSpaceMaskBoundaryCoords(refAnatT1)
#overlayBoundCoords=subjectSpaceMaskBoundaryCoords(overlayNifti)
#find which input has the highest resolution
refAnatT1Resolution=refAnatT1.header.get_zooms()
overlayResolution=overlayNifti.header.get_zooms()
#change datatype so that nilearn doesn't mess up
#it doesn't like bool so we have to do something about that
#somehow the data can be set to nifti, but can't be extracted as nifti, so...
overlayNifti=nib.nifti1.Nifti1Image(overlayNifti.get_data(),overlayNifti.affine,overlayNifti.header)
overlayNifti.set_data_dtype(overlayNifti.get_data().dtype)
#resample to the best resolution
#but also assuming that the refernce anatomy is ultimately the shape that we want
#this is going to cause huge problems if the user passes in a super cropped overlay
if np.prod(refAnatT1Resolution)>np.prod(overlayResolution):
print('resampling reference anatomy to overlay')
refAnatT1=resample_img(refAnatT1,target_affine=overlayNifti.affine[0:3,0:3])
overlayNifti=resample_img(overlayNifti,target_affine=overlayNifti.affine[0:3,0:3])
else:
print('resampling overlay to reference anatomy')
refAnatT1=resample_img(refAnatT1,target_affine=refAnatT1.affine[0:3,0:3])
overlayNifti=resample_img(overlayNifti,target_affine=refAnatT1.affine[0:3,0:3])
#crop the anatomy back down in case it has gotten overly widened
refAnatT1=crop_img(refAnatT1)
#now crop the overlay to the dimensions of the reference anatomy
overlayNifti=resample_img(overlayNifti,target_affine=refAnatT1.affine, target_shape=refAnatT1.shape)
#WARNING THIS INTRODUCES NIQUEST SAMPLING ERRORS INTO THE OVERLAY
dimsList=['x','y','z']
#brute force with matplotlib
for iDims in list(range(len(refAnatT1.shape))):
#this assumes that get_zooms returns zooms in subject space and not image space orientation
# which may not be a good assumption if the orientation is weird
#pick whichever input has the best resolution in this dimension
if refAnatT1Resolution[iDims]<=overlayResolution[iDims]:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],refAnatT1Resolution[iDims])
else:
subjectSpaceSlices=np.arange(convertedBoundCoords[0,iDims],convertedBoundCoords[1,iDims],overlayResolution[iDims])
#get the desired broadcast shape and delete current dim value
broadcastShape=list(refAnatT1.shape)
del broadcastShape[iDims]
#iterate across slices
for iSlices in list(range(len(subjectSpaceSlices))):
#set the slice list entry to the appropriate singular value
#THE SOLUTION WAS SO OBVIOUS. DONT USE A SINGLE SLICE FOR BOTH THE
#REFERNCE AND THE OVERLAY. DUH!
#actually this doesn't matter if we resample
currentRefSlice=wmaPyTools.roiTools.makePlanarROI(refAnatT1, subjectSpaceSlices[iSlices], dimsList[iDims])
#could be an issue if overlay nifti is signifigantly smaller
currentOverlaySlice=wmaPyTools.roiTools.makePlanarROI(overlayNifti, subjectSpaceSlices[iSlices], dimsList[iDims])
#set up the figure
fig,ax = plt.subplots()
ax.axis('off')
#kind of overwhelming to do this in one line
refData=np.rot90(np.reshape(refAnatT1.get_data()[currentRefSlice.get_data().astype(bool)],broadcastShape),1)
plt.imshow(refData, cmap='gray', interpolation='gaussian')
#kind of overwhelming to do this in one line
overlayData=np.rot90(np.reshape(overlayNifti.get_data()[currentOverlaySlice.get_data().astype(bool)],broadcastShape),1)
#lets mask out the background
[unique, counts] = np.unique(overlayData, return_counts=True)
backgroundVal=unique[np.where(np.max(counts)==counts)[0]]
plt.imshow(np.ma.masked_where(overlayData==backgroundVal,overlayData), cmap='jet', alpha=.75, interpolation='gaussian',vmin=0,vmax=np.nanmax(overlayNifti.get_data()))
curFig=plt.gcf()
cbaxes = inset_axes(curFig.gca(), width="5%", height="80%", loc=5)
plt.colorbar(cax=cbaxes, ticks=[0.,np.nanmax(overlayNifti.get_data())], orientation='vertical')
curFig.gca().yaxis.set_ticks_position('left')
curFig.gca().tick_params( colors='white')
# we use *2 in order to afford room for the subsequent blended images
figName='dim_' + str(iDims) +'_'+ str(iSlices*2).zfill(3)+'.png'
plt.savefig(figName,bbox_inches='tight',pad_inches=0.0)
#plt.clf()
plt.close()
import os
from PIL import Image
from glob import glob
#create blened images to smooth transitions between slices
if blendOption:
for iDims in list(range(len(refAnatT1.shape))):
dimStem='dim_' + str(iDims)
imageList=sorted(glob(dimStem+'*.png'))
for iImages in list(range(len(imageList)-1)):
thisImage=Image.open(imageList[iImages])
nextImage=Image.open(imageList[iImages+1])
blendedImage = Image.blend(thisImage, nextImage, alpha=0.5)
# 1 + 2 * iImages fills in the name space we left earlier
figName='dim_' + str(iDims) +'_'+ str(1+iImages*2).zfill(3)+'.png'
blendedImage.save(figName,'png')
for iDims in list(range(len(refAnatT1.shape))):
dimStem='dim_' + str(iDims)
img, *imgs = [Image.open(f) for f in sorted(glob(dimStem+'*.png'))]
img.save(os.path.join(saveDir,dimStem+'.gif'), format='GIF', append_images=imgs,
save_all=True, duration=1, loop=0)
plt.close('all')
[os.remove(ipaths) for ipaths in sorted(glob(dimStem+'*.png'))]
# #stop code
# if wmaPyTools.genUtils.is_docker():
# #borrowing from
# #https://github.com/brainlife/app-wmc_figures/blob/76c4cf6448a72299f2d70195f9177b75e3310934/main.py#L32-L38
# vdisplay.stop()
def densityGifsOfTract(tractStreamlines,referenceAnatomy,saveDir,tractName):
import os
import nibabel as nib
from glob import glob
import numpy as np
import dipy
import dipy.tracking.utils as ut
tractMask=ut.density_map(tractStreamlines, referenceAnatomy.affine, referenceAnatomy.shape)
densityNifti = nib.nifti1.Nifti1Image(tractMask, referenceAnatomy.affine, referenceAnatomy.header)
#now make the niftiGifs
crossSectionGIFsFromNifti(densityNifti,referenceAnatomy,saveDir)
filesToRename=[os.path.join(saveDir,'dim_'+ str(iDim) +'.gif') for iDim in range(3)]
for iFiles in filesToRename:
[path, file]=os.path.split(iFiles)
os.rename(iFiles,os.path.join(path,tractName+'_'+file))
#def densityGifsOfOverlay(tractStreamlines,referenceAnatomy,saveDir,tractName):
def radialTractEndpointFingerprintPlot(tractStreamlines,atlas,atlasLookupTable,tractName='tract',forcePlotLabels=None,saveDir=None,color=False):
"""radialTractEndpointFingerprintPlot(tractStreamlines,atlas,atlasLookupTable,tractName=None,saveDir=None)
A function used to generate radial fingerprint plots for input tracts, as
found in Folloni 2021. Plots absolute streamline count
Parameters
----------
tractStreamlines : streamlines type
Streamlines corresponding to the tract of interest
atlas : Nifti, int based
A nifti atlas that will be used to determine the endpoint connectivity
atlasLookupTable : pandas dataframe or file loadable to pandas dataframe
A dataframe of the atlas lookup table which includes the labels featured
in the atlas and their identities. These identities will be used
to label the periphery of the radial plot.
tractName : string, optional
The name of the tract to be used as a label in the output figure. No
input will result in there being no corresponding label. The default is None.
forcePlotLabels: array-like / list of int, optional
A list or array of integer values corresponding to labels in the input
atlas / atlasLookupTable that will be included in the output plot
REGARDLESS of whether streamlines connect to this region or not (e.g.
this option will force the plotting of 0 values for specificed labels)
saveDir : TYPE, optional
The directory in which to save the resultant radial plot. No input
will result in no save. The default is None.
Returns
-------
The figure generated
"""
import os
import matplotlib.pyplot as plt
import wmaPyTools.analysisTools
import pandas as pd
#just use this to get the column names, you can't be sure that all the names
#are there
[renumberedAtlasNifti,reducedLookUpTable]=wmaPyTools.analysisTools.reduceAtlasAndLookupTable(atlas,atlasLookupTable,removeAbsentLabels=False)
columnNames=reducedLookUpTable.columns
#don't pass the output of this to the next part, because the label numbers no longer match the atlas
[endpointsDF1, endpointsDF2]=wmaPyTools.analysisTools.quantifyTractEndpoints(tractStreamlines,atlas,atlasLookupTable)
if color==True:
endpointsDF1['color']=''
endpointsDF2['color']=''
streamColors, colorLut=colorStreamEndpointsFromParc(tractStreamlines,atlas)
#place label names in the color lut
colorLut['name']=''
for iRows,iLabels in enumerate(colorLut['labels'].to_list()):
colorLut['name'].iloc[iRows]=reducedLookUpTable.iloc[(reducedLookUpTable.iloc[:,0]==iLabels).to_numpy(),1].values[0]
#now do it the other way for the label values
for iRows,iNames in enumerate(colorLut['name'].to_list()):
if iNames in endpointsDF1['labelNames'].to_list():
endpointsDF1['color'].iloc[(endpointsDF1['labelNames']==iNames).to_numpy()]=pd.Series([colorLut['rgb_Color'].iloc[iRows]])
if iNames in endpointsDF2['labelNames'].to_list():
endpointsDF2['color'].iloc[(endpointsDF2['labelNames']==iNames).to_numpy()]=pd.Series([colorLut['rgb_Color'].iloc[iRows]])
#get the requested labels, if any
if not forcePlotLabels==None:
#get the sub table for the requeseted labels
forceTable=atlasLookupTable[atlasLookupTable[columnNames[0]].isin(forcePlotLabels)]
#find the labels that aren't already there
missingLabels1=forceTable[columnNames[1]][~forceTable[columnNames[1]].isin(endpointsDF1['labelNames'])]
missingLabels2=forceTable[columnNames[1]][~forceTable[columnNames[1]].isin(endpointsDF2['labelNames'])]
#now append them to the tables
endpointsDF1=endpointsDF1.append(pd.DataFrame(missingLabels1,columns=['labelNames']),ignore_index=True)
endpointsDF2=endpointsDF2.append(pd.DataFrame(missingLabels2,columns=['labelNames']),ignore_index=True)
#now set the nans to zero
endpointsDF1['endpointCounts']= endpointsDF1['endpointCounts'].fillna(0)
endpointsDF2['endpointCounts']= endpointsDF2['endpointCounts'].fillna(0)
if color==True:
figure1=basicRadarPlot(list(endpointsDF1['labelNames']),list(endpointsDF1['endpointCounts']),COLORS=endpointsDF1['color'].to_list())
figure2=basicRadarPlot(list(endpointsDF2['labelNames']),list(endpointsDF2['endpointCounts']),COLORS=endpointsDF2['color'].to_list())
else:
figure1=basicRadarPlot(list(endpointsDF1['labelNames']),list(endpointsDF1['endpointCounts']))
figure2=basicRadarPlot(list(endpointsDF2['labelNames']),list(endpointsDF2['endpointCounts']))
figure1.get_axes()[0].set_title(tractName+'\nRAS endpoints')
figure1.get_axes()[0].set_facecolor([0,0,1,.15])
#figure1.patch.set_facecolor([0,0,1,.2])
figure2.get_axes()[0].set_title(tractName+'\nLPI endpoints')
figure2.get_axes()[0].set_facecolor([1,0,0,.15])
#figure2.patch.set_facecolor([1,0,0,.2])
if saveDir==None:
saveDir=os.getcwd()
figure1.savefig(os.path.join(saveDir,tractName+'_RAS_endpointMap.eps'))
figure2.savefig(os.path.join(saveDir,tractName+'_LPI_endpointMap.eps'))
def radialTractEndpointFingerprintPlot_MultiSubj(tractTractogramList,atlasList,atlasLookupTable,tractName='tract',forcePlotLabels=None,saveDir=None):
"""
A function used to generate radial fingerprint plots for input tracts, as
found in Folloni 2021. Plots proportional streamline count, averaged
across input variants, with error bars. Secret Trick: if you input two
instances of the same tract/atlas (as opposed to inputs for different
subjects), will simply plot the proportion for that single subject.
Parameters
----------
tractTractogramList : list of streamlines type
Streamlines corresponding to the tract of interest
atlasList: list of Nifti, int based
A nifti atlas that will be used to determine the endpoint connectivity
atlasLookupTable : pandas dataframe or file loadable to pandas dataframe
A dataframe of the atlas lookup table which includes the labels featured
in the atlas and their identities. These identities will be used
to label the periphery of the radial plot.
tractName : string, optional
The name of the tract to be used as a label in the output figure. No
input will result in there being no corresponding label. The default is None.
forcePlotLabels: array-like / list of int, optional
A list or array of integer values corresponding to labels in the input
atlas / atlasLookupTable that will be included in the output plot
REGARDLESS of whether streamlines connect to this region or not (e.g.
this option will force the plotting of 0 values for specificed labels)
saveDir : TYPE, optional
The directory in which to save the resultant radial plot. No input
will result in no save. The default is None.
Returns
-------
None.
"""
import pandas as pd
import numpy as np
import os
from matplotlib import pyplot as plt
import wmaPyTools.analysisTools
#just use this to get the column names, you can't be sure that all the names
#are there
[renumberedAtlasNifti,reducedLookUpTable]=wmaPyTools.analysisTools.reduceAtlasAndLookupTable(atlasList[0],atlasLookupTable,removeAbsentLabels=False)
columnNames=reducedLookUpTable.columns
RASendpointData=[]
LPIendpointData=[]
for iTracts in range(len(tractTractogramList)):
[currentRAS,currentLPI]=wmaPyTools.analysisTools.quantifyTractEndpoints(tractTractogramList[iTracts],atlasList[iTracts],atlasLookupTable)
RASendpointData.append(currentRAS)
LPIendpointData.append(currentLPI)
#normalize them
for iTracts in range(len(tractTractogramList)):
RASendpointData[iTracts]['endpointCounts']=RASendpointData[iTracts]['endpointCounts'].divide(RASendpointData[iTracts]['endpointCounts'].sum())
LPIendpointData[iTracts]['endpointCounts']=LPIendpointData[iTracts]['endpointCounts'].divide(LPIendpointData[iTracts]['endpointCounts'].sum())
firstRASDF=RASendpointData[0]
firstLPIDF=LPIendpointData[0]
for iTracts in range(1,len(tractTractogramList)):
firstRASDF=pd.merge(firstRASDF,RASendpointData[iTracts], on='labelNames', how='outer')
firstLPIDF=pd.merge(firstLPIDF,LPIendpointData[iTracts], on='labelNames', how='outer')
#set NaNs to 0
firstRASDF=firstRASDF.fillna(0)
firstLPIDF=firstLPIDF.fillna(0)
#compute means and variances
firstLPIDF[['meanProportion','proportionSTD']]=pd.DataFrame(np.atleast_2d(np.squeeze(np.asarray([np.mean(firstLPIDF.iloc[:,1:-1],axis=1),np.std(firstLPIDF.iloc[:,1:-1],axis=1)])).T), index=firstLPIDF.index)