def refreshGUI(self):
while (True):
self.clear()
ev = evaluateFile()
similarCharters, sameCategory = ev.getOpenFile()
# similarCharters = ["Project Charter_DivA2.docx"]
# sameCategory = []
self.printGUI(similarCharters, sameCategory)
self.notify(title = 'Big Brother',
subtitle = 'Similar Project Charters',
message = 'Found project charters similar to what you are working on! Click here to find out more!',
execute = 'python showDoc.py',
activate = 'com.apple.Terminal')
sleep(100)
from processFile import processFile
# gt file
hdf5_gt_file = '/groups/turaga/home/turagas/data/FlyEM/fibsem_medulla_7col/tstvol-520-2-h5/groundtruth_seg_thick.h5' #groundtruth_aff.h5
# input models
model_base_folder = '/groups/turaga/home/turagas/research/caffe_v1/pygt_models/fibsem'
fibsemFolders = ['2','3','4','5','6']
iters = [30000,70000]
# output folders
train = False # which dataset to evaluate
h5OutputFilenames = ["data_tier2/test/output_"+str(iters[j])+"/"+"tstvol-2_"+fibsemFolders[i] for j in range(len(iters)) for i in range(len(fibsemFolders))]
randOutputFolder = ['data_tier2/test/out/fibsem' +fibsemFolders[i]+ '_'+str(iters[j])+'/' for j in range(len(iters)) for i in range(len(fibsemFolders))]
# settings
threshes = [i*2000 for i in range(1,6)]+[i*20000 for i in range(2,16)] # default: 100...1,000...100,000
funcs = ['square'] #'linear','threshold','watershed','lowhigh'
save_segs = False
for iter_idx in range(len(iters)):
for fibsem_idx in range(len(fibsemFolders)):
start = time.clock()
processFile(model_base_folder+fibsemFolders[fibsem_idx]+'/',iters[iter_idx],h5OutputFilenames[iter_idx*len(fibsemFolders)+fibsem_idx],train)
evaluateFile([hdf5_gt_file,h5OutputFilenames[fibsem_idx]+'.h5',threshes,funcs,save_segs,randOutputFolder[iter_idx*len(fibsemFolders)+fibsem_idx]])
#averageAndEvaluateFiles([hdf5_gt_file,h5OutputFilenames, threshes,funcs,save_segs,randOutputFolder[iter_idx*len(fibsemFolders)+fibsem_idx]]) # for averaging
print("time elapsed ",time.clock()-start)
iters = [10000*i for i in range(14,21)] hdf5_gt_file = '/groups/turaga/home/turagas/data/FlyEM/fibsem_medulla_7col/tstvol-520-'+vol+'-h5/groundtruth_seg_thick.h5' #groundtruth_aff.h5 # settings threshes = [i*2000 for i in range(1,6)]+[i*20000 for i in range(2,16)] # default: 100...1,000...100,000 funcs = ['square'] #'linear','threshold','watershed','lowhigh' save_threshes = [] #threshes process = True eval = True # output folders h5OutputFilenames = ["data_tier2/"+t+"/output_"+str(iters[j])+"/"+"tstvol-"+vol+"_"+fibsemFolders[i] for j in range(len(iters)) for i in range(len(fibsemFolders))] randOutputFolder = ['data_tier2/'+t+'/out/fibsem' +fibsemFolders[i]+ '_'+str(iters[j])+'/' for j in range(len(iters)) for i in range(len(fibsemFolders))] for iter_idx in range(len(iters)): for fibsem_idx in range(len(fibsemFolders)): if process: processFile(model_base_folder+fibsemFolders[fibsem_idx]+'/',iters[iter_idx],h5OutputFilenames[iter_idx*len(fibsemFolders)+fibsem_idx],train) if eval: evaluateFile([hdf5_gt_file,h5OutputFilenames[iter_idx*len(fibsemFolders)+fibsem_idx]+'.h5',threshes,funcs,save_threshes,randOutputFolder[iter_idx*len(fibsemFolders)+fibsem_idx]]) # this part might not work h5_filenames_to_average = ["/tier2/turaga/singhc/"+t+"/output_"+str(iters[iter_idx])+"/"+"tstvol-"+vol+"_"+fibsemFolders[i] for i in range(len(fibsemFolders))] out_folder = '/tier2/turaga/singhc/'+t+'/out/fibsemave_'+str(iters[iter_idx]) if process: averageFiles(h5_filenames_to_average,'/tier2/turaga/singhc/'+t+'/output_'+str(iters[iter_idx])+'/tstvol-'+vol+'_ave.h5') if eval: evaluateFile([hdf5_gt_file,'/tier2/turaga/singhc/'+t+'/output_'+str(iters[iter_idx])+'/tstvol-'+vol+'_ave.h5',threshes,funcs,save_threshes,out_folder+'/'])
