R8206/loggerF1.py
13c87b6c2703master

loggerF1.py
View Options

			import os
			import torch
			import numpy as np
			from .f1 import PRFromHistograms, F1FromPR

			def reverse(t):
			idx = [i for i in range(t.size(0)-1, -1, -1)]
			idx = torch.LongTensor(idx)
			it = t.index_select(0, idx)
			return it

			class LoggerF1:

			def __init__(self,logdir,fname,transform,nBins=10000,saveBest=False):
			self.log_dir=logdir
			self.name=fname
			self.log_file=os.path.join(self.log_dir,"logF1"+self.name+".txt")
			text_file = open(self.log_file, "w")
			text_file.close()
			self.preproc=transform
			self.nBins=nBins
			self.hPos=torch.zeros(self.nBins)
			self.hNeg=torch.zeros(self.nBins)
			self.bestF1=0
			self.saveBest=saveBest

			def add(self,l,output,target):
			o,t=self.preproc(output.cpu().data,target.cpu().data)
			pos=o[t==1]
			neg=o[t==0]
			self.hPos+=pos.histc(self.nBins,0,1)
			self.hNeg+=neg.histc(self.nBins,0,1)

			def logEpoch(self,net):
			precision,recall=PRFromHistograms(self.hPos,self.hNeg)
			f1s=F1FromPR(precision,recall)
			f=f1s.max()
			text_file=open(self.log_file, "a")
			text_file.write('{}\n'.format(f))
			text_file.close()
			if self.saveBest and f > self.bestF1:
			self.bestF1=f
			torch.save({'state_dict': net.state_dict()},
			os.path.join(self.log_dir,
			'net_'+self.name+'_bestF1.pth'))


			self.hPos.zero_()
			self.hNeg.zero_()