to_hierarchical_dataframe() DataFrames are too wide and long

[lvsltz]

Not sure if in your control but I mention this anyway.
to_hierarchical_dataframe() and to_denormalized_dataframe() generate very wide and long DataFrames. To save some space and facilitate visualization, the stimulus, response and electrode columns from intracellular_recordings should contain fewer details.

[oruebel]

@lvsltz that makes sense. I think the proper solution would be to wrap the stimulus and response columns in a separate class (and to give them a new neurodata_type) so that we can control the representation. That is certainly possible but requires a bit of coding and changes to schema, i.e., it may take a bit before I can get to this issue.

[oruebel]

A simple workaround would be to set the maximum width for columns in the pandas display, e.g.,:

import pandas
pandas.set_option("display.max_colwidth", 30)

This allows you to control the column width in the rendering. However, it would still be nice to have nicer text representations for the stimulus and response columns.

[lvsltz]

Yes, this can work too. What I was thinking of is more to decorate python's display() or hdmf's Container when a denormalized_dataframe is shown, like below:

from pynwb.icephys import PatchClampSeries, IntracellularElectrode

def newrepr(self):
    cls = self.__class__
    if cls.__name__ in ['VoltageClampStimulusSeries',
                        'VoltageClampSeries',
                        'CurrentClampStimulusSeries',
                        'CurrentClampSeries',
                        'IntracellularElectrode']:
        template = "%s" % (self.name)
    else:
        template = self.__repr__()
    return template

IntracellularElectrode.__repr__ = newrepr
PatchClampSeries.__repr__ = newrepr

But this issue is really not very important, you can even close it :)