df.plot with two plots in the same axes and 2nd one is line plot makes different chart between 0.23.4 and 0.25.0

[wh243s]

Code Sample, a copy-pastable example if possible

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('iris.csv')
dfg = df.groupby('species').mean()
dfg['swratio'] = dfg.sepal_width / dfg.sepal_length

fig, ax = plt.subplots(figsize=(10, 6))
dfg[['sepal_length', 'sepal_width', 'petal_length', 'petal_width']].plot.bar(ax=ax, rot=0, stacked=False)
dfg[['swratio']].plot.line(label='Sepal Width Ratio', 
                 ax=ax, secondary_y=True, mark_right=False, 
                 rot=0, marker='o', markersize=3, color='c')
ax.grid(True)
pngfile = 'Output.png'
fig = plt.gcf()
fig.savefig(pngfile, bbox_inches = 'tight', dpi=100)

Problem description

dataframe.plot with secondary_y=True makes different chart between 0.23.4 and 0.25.0 as attached.

Note: We receive a lot of issues on our GitHub tracker, so it is very possible that your issue has been posted before. Please check first before submitting so that we do not have to handle and close duplicates!

Note: Many problems can be resolved by simply upgrading pandas to the latest version. Before submitting, please check if that solution works for you. If possible, you may want to check if master addresses this issue, but that is not necessary.

For documentation-related issues, you can check the latest versions of the docs on master here:

https://pandas-docs.github.io/pandas-docs-travis/

If the issue has not been resolved there, go ahead and file it in the issue tracker.

Expected Output

Output of pd.show_versions()

[paste the output of pd.show_versions() here below this line]

[TomAugspurger]

@wh243s can you debug to figure out what is going on?

[wh243s]

I've tried three different parameters on 0.25.0.

1. dfg[['swratio']].plot.line(ax=ax, secondary_y=True)
2. dfg[['swratio']].plot.line(ax=ax)
3. dfg[['swratio']].plot.line()

The attached are the results and csv being used.

iris _renamed_from_csv.txt

[wh243s]

3. dfg[['swratio']].plot.line() has two chart in png and only the last one shows. The attached is the first one.
   

[wh243s]

When line and bar charts are switched, it looks fine.

import pandas as pd
import matplotlib.pyplot as plt

df = pd.read_csv('iris.csv')
dfg = df.groupby('species').mean()
dfg['swratio'] = dfg.sepal_width / dfg.sepal_length

fig, ax = plt.subplots(figsize=(10, 6))
dfg[['swratio']].plot.line(ax=ax, color='c', rot=0)
dfg[['sepal_length', 'sepal_width', 'petal_length', 'petal_width']].plot.bar(ax=ax, secondary_y=True, rot=0)
ax.grid(True)
plt.title('plot.bar(ax=ax, secondary_y=True, rot=0)')
pngfile = 'GeneratedFromPahdas0.25.0_switched_line_and_bar.png'
fig = plt.gcf()
fig.savefig(pngfile, bbox_inches = 'tight', dpi=100)

[wh243s]

It happens when the second chart is the line chart in the same axes.

[arthurire]

@TomAugspurger https://github.com/pandas-dev/pandas/issues/27796. Same issue. This also happens when I use seaborn to plot a heatmap.

[TomAugspurger]

You're saying #27796 is a duplicate of this?

[arthurire]

You're saying #27796 is a duplicate of this?

I guess so. I mean the results(plots) we get are the same: line charts would re-define the range of x-axis, which makes the line chart starts from the origin. But there should be a certain distance between the line chart and the origin.

[TomAugspurger]

Ok I'll close that one then.

…

On Thu, Aug 8, 2019 at 7:17 AM arthurire ***@***.***> wrote: You're saying #27796 <#27796> is a duplicate of this? I guess so. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#27686?email_source=notifications&email_token=AAKAOISFT34CCORT7VDIO6DQDQFDZA5CNFSM4IINZVVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD33NPZA#issuecomment-519493604>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAKAOIXYBGAIYY6ADQM5A7LQDQFDZANCNFSM4IINZVVA> .

[ryanrolds]

The fix to this causes broader limits that were set prior to calls to pandas.DataFrame.plot.(line|area) to be lost. The fix is to manually set the desired axis limit after the calls to DataFrame.plot, but it was still a bit of a surprise to have it narrow the x-axis after the was specifically set larger so that projection lines (pyplot.axvline) would be visible.