You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When I attempt to add a block of signals larger than about 2000000 samples (for 256 channels) to a BinWriter object, I get this error:
---------------------------------------------------------------------------
error Traceback (most recent call last)
<ipython-input-7-6fddd83ebbc4> in <module>
1 B = BinWriter(sampling_rate=sr)
----> 2 B.add_block(signals.astype(np.float32))
3 W = Writer(join('examples', 'sample.mff'))
4 startdatetime = datetime.strptime('1984-02-18T14:00:10.000000+0100',
5 "%Y-%m-%dT%H:%M:%S.%f%z")
~/Repositories/mffpy/mffpy/bin_writer.py in add_block(self, data, offset_us)
99 )
100 # Write header/data to stream, and add an epochs block
--> 101 write_header_block(self.stream, self.header)
102 self.append(data.tobytes())
103 self._add_block_to_epochs(num_samples, offset_us=offset_us)
~/Repositories/mffpy/mffpy/header_block.py in write_header_block(fp, hdr)
130 """write HeaderBlock `hdr` to file pointer `fp`"""
131 fp.write(struct.pack('4i',
--> 132 1, hdr.header_size, hdr.block_size, hdr.num_channels))
133 num_samples = (hdr.block_size//hdr.num_channels) // 4
134 # Write channel offset into the data block
error: 'i' format requires -2147483648 <= number <= 2147483647
I am able to work around this error for now by splitting the signals into multiple block and adding the blocks one at a time with no gap in time between them. I run into this error often enough that it would be good to look into how we could allow adding of larger blocks of EEG to a BinWriter object.
The text was updated successfully, but these errors were encountered:
@jusjusjus I think this is still worth solving. Basically we can't write a block of data of a certain size. I think we should at least throw out a message stating the maximum size of a block that we can write when this error occurs.
When I attempt to add a block of signals larger than about 2000000 samples (for 256 channels) to a BinWriter object, I get this error:
I am able to work around this error for now by splitting the signals into multiple block and adding the blocks one at a time with no gap in time between them. I run into this error often enough that it would be good to look into how we could allow adding of larger blocks of EEG to a BinWriter object.
The text was updated successfully, but these errors were encountered: