When saving a spiketrain with a waveform array of shape (0,0,0) and loading it again using the NixIO, the new shape of the waveform array is (0,).
Here is some code showing this explicitely:
from neo import Block, Segment, SpikeTrain, NixIO
import quantities as pq
import numpy as np
bl = Block()
seg = Segment()
bl.segments.append(seg)
st = SpikeTrain([1,2,3]*pq.s,t_start=0*pq.s,t_stop=10*pq.s,
sampling_rate=1*pq.Hz)
wfs = np.array([]).reshape((0,0,0))*pq.V
st.waveforms = wfs
bl.segments[0].spiketrains.append(st)
with NixIO('test.nix','ow') as nio:
nio.write_block(bl)
print('waveform shape before writing {}'
''.format(bl.segments[0].spiketrains[0].waveforms.shape))
with NixIO('test.nix','ro') as nio:
block = nio.read_block()
print('waveform shape after loading {}'
''.format(block.segments[0].spiketrains[0].waveforms.shape))
The prints are:
waveform shape before writing (0, 0, 0)
waveform shape after loading (0,)
When saving a spiketrain with a waveform array of shape
(0,0,0)and loading it again using the NixIO, the new shape of the waveform array is(0,).Here is some code showing this explicitely:
The prints are: