data_reader.py 文件源码

def next_batch(self, batch_size = 50) :

            start = self._index_in_epoch
            if ( self._epochs_completed == 0 ) and ( start == 0 ) :
                self.batch_size = batch_size
                while np.modf(float(self._ndata)/self.batch_size)[0] > 0.0 :
                     print 'Warning! Number of data/ batch size must be an integer.'
                     print 'number of data: %d' % self._ndata
                     print 'batch size: %d'     % self.batch_size
                     self.batch_size = int(input('Input new batch size: '))
                print 'batch size : %d'    % self.batch_size
                print 'number of data: %d' % self._ndata

            self._index_in_epoch += self.batch_size
            if self._index_in_epoch > self._ndata :
                # Number of training epochs completed
                self._epochs_completed += 1
                # Shuffle data
                random.shuffle(self.shuffle_index)
                self._images = self._images[self.shuffle_index]
                self._labels = self._labels[self.shuffle_index]
                # Reinitialize conunter
                start = 0
                self._index_in_epoch = self.batch_size
                assert self.batch_size <= self._ndata
            end = self._index_in_epoch
            return self._images[start:end], self._labels[start:end]