0ac2bf1c2cafee4cc9555c9f09a17143b05b1a88,brian2/synapses/spikequeue.py,SpikeQueue,prepare,#SpikeQueue#Any#Any#Any#,91

Before Change

        max_events = 0
        synapse_sources = synapse_sources[:]
        for source in xrange(self._source_end - self._source_start):
            indices = np.flatnonzero(synapse_sources == (source +
                                                         self._source_start))
            size = len(indices)
            self._neurons_to_synapses.append(indices)
            max_events = max(max_events, size)

        n_steps = max_delays + 1
        
        // Adjust the maximum delay and number of events per timestep if necessary
        // Check if delays are homogeneous

After Change

        // Prepare the data structure used in propagation
        synapse_sources = synapse_sources[:]        
        nsource = self._source_end - self._source_start
        ss = np.ravel(synapse_sources)
        I = np.argsort(ss)
        ss_sorted = ss[I]
        splitinds = np.searchsorted(ss_sorted, np.arange(self._source_start, self. _source_end+1))
        self._neurons_to_synapses = [I[splitinds[j]:splitinds[j+1]] for j in xrange(len(splitinds)-1)]
        max_events = max(map(len, self._neurons_to_synapses))

        n_steps = max_delays + 1
        
        // Adjust the maximum delay and number of events per timestep if necessary

Instances