fig.append_trace() raises ValueError if row arg is not type int.

[mark-datanaltek]

fig.append_trace() raises ValueError if the row arg is not type int, even if arg is int-like:
DEBUG row:1, type(row):<class 'numpy.int64'>

If I explicitly cast to int with row=int(row), all is well. The issue may be due to the isinstance check at line 1536 of basedatatypes.py?

/usr/local/lib/python3.7/site-packages/plotly/basedatatypes.py in _validate_rows_cols(name, n, vals)
1535
1536 if [r for r in vals if not isinstance(r, int)]:
-> 1537 BaseFigure._raise_invalid_rows_cols(name=name, n=n, invalid=vals)
1538 else:
1539 BaseFigure._raise_invalid_rows_cols(name=name, n=n, invalid=vals)

/usr/local/lib/python3.7/site-packages/plotly/basedatatypes.py in _raise_invalid_rows_cols(name, n, invalid)
1524 )
1525
-> 1526 raise ValueError(rows_err_msg)
1527
1528 @staticmethod

ValueError:
If specified, the rows parameter must be a list or tuple of integers
of length 1 (The number of traces being added)

    Received: [1]

I am working with the latest version:

import plotly
plotly.version
'4.6.0'

[emmanuelle]

Thank you @mark-datanaltek this is indeed a bug. To reproduce

import numpy as np
import plotly.graph_objects as go
indices_rows = np.array([0], dtype=np.int)
indices_cols = np.array([0], dtype=np.int)
fig = go.Figure()
fig.add_trace(go.Scatter(y=[1]), row=indices_rows[0], col=indices_cols[0])
fig.show()

[emmanuelle]

This could be easily fixed by modifying slightly this line of code https://github.com/plotly/plotly.py/blob/master/packages/python/plotly/plotly/basedatatypes.py#L1531
Since numpy is not a dependency we would have to check whether numpy is imported

try:
    import numpy as np
    int_type = (int, np.integer)
except ImportError:
    int_type = (int,)
....
if [r for r in vals if not isinstance(r, int_type)]:
....

[MCBoarder289]

Thanks @emmanuelle for tagging this as a good first issue. I'll see if I can give it a try here!

[MCBoarder289]

Ran into an exception with the reproducing example once I got this working:

Exception: In order to reference traces by row and column, you must first use plotly.tools.make_subplots to create the figure with a subplot grid.

Here is one that works with a fix in place. Seeing if I can write this into a test before creating a pull request:

import numpy as np
import plotly.graph_objects as go
from plotly.subplots import make_subplots
indices_rows = np.array([1], dtype=np.int)
indices_cols = np.array([1], dtype=np.int)
fig = make_subplots(rows=1, cols=1)
fig.add_trace(go.Scatter(y=[1]), row=indices_rows[0], col=indices_cols[0])

[MCBoarder289]

I have put a pull request up here: #2451
Open to any feedback on this!

[MCBoarder289]

@emmanuelle Thanks for the review! Should we link #2451 now that's been approved and merged for this ticket? I wasn't sure how to do that when I made the PR.

[emmanuelle]

@MCBoarder289 you can write "closes #xx" with the issue number when you start a PR, then the issue will be closed automatically on merging the PR.

Closing since this has been closed by #2451.