Plotly.py: plotly.tools.make_subplots() breaks down if rows>=100

[oiegorov]

For some reason all data in subplots get jumbled up if 'rows' is 100 or higher. It works fine for rows=99. Here is an example code:

N = 100
v_space = 0

fig = tools.make_subplots(
    rows=N,
    cols=1,
    vertical_spacing=v_space
)

for i in range(N):
    x1 = 1
    y1 = 1
    trace = go.Scatter(
        x=[x1],
        y=[y1],
    )
    fig.append_trace(trace, i + 1, 1)

fig['layout'].update(height=N * 200)
py.iplot(fig)

If vertical_spacing is not zero and fig['layout'].update(height=N*(200+vertical_spacing)), it breaks down for smaller values of 'rows'.

[jonmmease]

Thanks for the report @oiegorov,

I think this is the same issue that was brought up in the forums here: https://community.plot.ly/t/domain-issues-with-multiple-subplots/12282/2

With plotly.py version 3 we get a more informative error (instead of the jumbled result you were seeing)

ValueError: 
    Invalid value of type 'builtins.float' received for the 'domain[1]' property of layout.yaxis
        Received value: 1.0000000000000007

    The 'domain[1]' property is a number and may be specified as:
      - An int or float in the interval [0, 1]

Looks like a numerical precision error. We probably just need a min(1.0, val) somewhere.

[jonmmease]

Hi @oiegorov ,

I just pushed out release candidates for plotly.py 3.1.1 and plotlywidget 0.2.1. Installation instructions for the release candidates are at https://github.com/plotly/plotly.py/blob/bc1d4d188ab999cd9c21e3a7908729f156bc200f/README.md.

If you have a chance to confirm that this issue is resolved in the release candidate that would be awesome!

[Rimesh]

Hi @jonmmease ,

It looks like the issue is still not resolved.
make_subplots breaks down at rows > 70

Here is an example code:

from plotly.offline import plot
import plotly.graph_objs as go
from plotly import tools

X = list(range(50))
Y = list(range(50))

n_rows = 70 # make_subplots breaks down if rows > 70

fig = tools.make_subplots(rows=n_rows, cols=1)
fig['layout'].update(height=5000, width=1000)

for i in range(n_rows):
    trace = go.Scatter(x = X, y = Y)
    fig.append_trace(trace, i+1, 1)

plot(fig)

[jonmmease]

Hi @Rimesh, thanks for letting us know. Could you add the error message that you're seeing?

[jonmmease]

@Rimesh, I gave it a try and I don't see an error message, but I do see that the subplot y-axes start overlapping. If this is what you're seeing as well, could you open a new issue? I think this is a different underlying issue. Thanks!

[baasman]

I'm seeing this as well, using 3.5.0

[elanter]

Seeing this as well. using 4.2.1

[albertsugi-tallridge]

Hi @jonmmease , I am using plotly 4.5.1 and am getting this error too as i tried creating a subplot with 63 rows. it gives an error if the row is bigger than 58 rows.

Here is the code:
fig = make_subplots(rows=63, cols=1, shared_xaxes=True, vertical_spacing=0.02) size = len(asinhourlydict) xvalues = ["12am", "1am", "2am","3am","4am","5am","6am","7am","8am","9am","10am","11am","12pm","1pm", "2pm","3pm","4pm","5pm","6pm","7pm","8pm","9pm","10pm","11pm" ] for key,value in asinhourlydict.items(): fig.add_trace(go.Bar(x=xvalues,y=value),row=size,col= 1) fig.update_yaxes(title_text=key, row=size, col=1) size = size -1 fig.write_html('violationbarchart.html', config={ "displayModeBar" : False} )

and here is the error:

ValueError: Invalid value of type 'builtins.float' received for the 'domain[0]' property of layout.yaxis Received value: 1.0038095238095237 The 'domain[0]' property is a number and may be specified as: - An int or float in the interval [0, 1]

Is there a way to fix this? Or is it my code that have bugs? Thanks in advance!

[robroc]

I'm having this issue with Plotly Express in Plotly 4.7.1, when trying to create faceted subplots with 92 plots (at 6 per row it's 16 rows).

[nicolaskruchten]

@albertsugi-tallridge you're encountering bug https://github.com/plotly/plotly.py/issues/2556 and you can get around it by reducing your vertical_spacing

@robroc yours is more akin to https://github.com/plotly/plotly.py/issues/2026

[nicolaskruchten]

In Plotly Express you can now (as of version 4.9) avoid this problem by using facet_row_spacing and facet_col_spacing https://plotly.com/python/facet-plots/#controlling-facet-spacing