-1

I am using the seaborns displot() for plot a CDF distribution of my dataframe.

Here is how I plot the figures.

First figure:

g = sns.displot(data=df, x=df['trip_duration']/60, kind='ecdf',)
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
plt.title('Trip durations')

enter image description here

And the second figure as:

g = sns.displot(data=df, x=df['trip_duration']/60, kind='ecdf', hue='travelmode', palette=palette)
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
plt.title('Trip durations per mode')

enter image description here

These figures are plotted separately. I would like to have them in a (1row, 2cols) format, for easy comparison. Plus, that way the would have the same x-axis scale. Right now, they have different scales even though I specify the range limit (0, 180)

edit

I tried this (first answer,)

fig, axs = plt.subplots(nrows=2, ncols=1, sharex=True)
# fig, axs = plt.subplots(nrows=1, ncols=2, sharey=True)
g = sns.displot(data=df, x=df['trip_duration']/60, kind='ecdf', ax=axs[0])
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
axs[0].set_title('Trip durations')

g = sns.displot(data=df, x=df['trip_duration']/60, kind='ecdf', hue='travelmode', palette=palette, ax=axs[1])
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
axs[1].set_title('Trip durations per mode')

enter image description here

arilwan
  • 3,374
  • 5
  • 26
  • 62

1 Answers1

1

First: you need to pass over matplotlib, a create subplots.

Second:There is a problem in your question if you want one row two columns you cannot share the same X values only Y values. If you want 2 rows 1 column is ok. I leave the options in the code

Once created you need to pass the ax where you want seaborn to plot the figure

fig, axs = plt.subplots(nrows=2, ncols=1, sharex=True)
# fig, axs = plt.subplots(nrows=1, ncols=2, sharey=True)
g = sns.ecdfplot(data=df, x=df['trip_duration']/60, ax=axs[0])
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
axs[0].set_title('Trip durations')

g = sns.ecdfplot(data=df, x=df['trip_duration']/60, hue='travelmode', palette=palette, ax=axs[1])
g.set(xlim=(0, 180))
g.set_axis_labels("Time (min)","CDF")
axs[1].set_title('Trip durations per mode')
Lucas M. Uriarte
  • 2,403
  • 5
  • 19
  • Doesn't work tthat way since `displot()` is a figure label. `UserWarning: displot is a figure-level function and does not accept the ax= paramter. You may wish to try ecdfplot. warnings.warn(msg, UserWarning)` – arilwan Sep 12 '22 at 11:12
  • Oh great! I can use `ecdfplot`. – arilwan Sep 12 '22 at 11:15