Display Pandas DataFrame

[white-gecko]

Is it possible to provide a pandas DataFrame? This should have the same structure as what the DataTable expects but somehow I don't have success.

data_table = DataTable(
    columns = [
        DataTableColumn("path"),
        DataTableColumn("count"),
    ],
    data=DataFrame(data=[["/", 1],["/home", 2],["/tmp",3]], columns=["path", "count"]),
    cell_selection=True,
    with_scrollbar=True,
)

But I get:

TypeError: 'str' object does not support item assignment

So far I'm doing some uncoordinated hacking, If this is a dump issue, just close it uncommented.

[tonycpsu]

One problem there is that the DataFrame name from Pandas collides with the one from panwid. Most people import Pandas as pd and then use pd.DataFrame.

I originally tried using Pandas as the underlying data structure for the table contents, but found that raccoon was a better fit for most of my use cases at the time. Pandas has changed a lot since then, so maybe it's worth investigating again or at least giving users the option of which to use, but I think the only way to do it now would be to use df.iterrows().

[tonycpsu]

A wrapper class like this is probably the best way to go for now:

#!/usr/bin/env python3

import urwid
from panwid.datatable import *
import numpy as np
import pandas as pd


def unhandled_input(key):
    if key in ("q", "Q"):
        raise urwid.ExitMainLoop()

class PandasDataTable(DataTable):

    def __init__(self, pdf, *args, **kwargs):

        self.pdf = pdf
        super().__init__(*args, **kwargs)

    @property
    def columns(self):
        return [
            DataTableColumn(c)
            for c in self.pdf.columns
        ]

    def query(self, *args, **kwargs):
        for i, row in self.pdf.iterrows():
            yield row


def main():

    pdf = pd.DataFrame(
        np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]]),
        columns=['a', 'b', 'c']
    )

    data_table = PandasDataTable(pdf)

    loop = urwid.MainLoop(
        urwid.Frame(data_table),
        unhandled_input=unhandled_input
    )
    loop.run()


if __name__ == "__main__":
    main()

[tonycpsu]

...which I just realized only works with recent changes I've made to the unstable branch, which I just pushed a moment ago.

[white-gecko]

I'm on the unstable branch anyways ;-) Thank you!