Develop Reference

Non-overlapping legend and axes (e.g. in matplotlib)

I need the plot legend to appear side-by-side to the plot axes, i.e. outside of the axes and non-overlapping.
The width of the axes and the legend should adjust "automatically" so that they both fill the figure w/o them to overlap or the legend to be cut, even when using tight layout. The legend should occupy a minor portion of the figure (let's say max to 1/3 of figure width so that the remaining 2/3 are dedicated to the actual plot).
Eventually, the font of the legend entries can automatically reduce to meet the requirements.
I've read a number of answers regarding legend and bbox_to_anchor in matplotlib with no luck, among which:
how to put the legend out of the plot
moving matplotlib legend outside of the axis makes it cutoff by the figure box
I tried by creating a dedicated axes in which to put the legend so that plt.tight_layout() would do its job properly but then the legend only takes a minor portion of the dedicated axes, with the result that a lot of space is wasted. Or if there isn't enough space (the figure is too small), the legend overlaps the first axes anyway.
import matplotlib.pyplot as plt
import numpy as np
# generate some data
x = np.arange(1, 100)
# create 2 side-by-side axes
fig, ax = plt.subplots(1,2)
# create a plot with a long legend
for ii in range(20):
ax[0].plot(x, x**2, label='20201110_120000')
ax[0].plot(x, x, label='20201104_110000')
# grab handles and labels from the first ax and pass it to the second
hl = ax[0].get_legend_handles_labels()
leg = ax[1].legend(*hl, ncol=2)
plt.tight_layout()
I'm open to use a package different from matplotlib.

Instead of trying to plot the legend in a separate axis, you can pass loc to legend:
# create 2 side-by-side axes
fig, ax = plt.subplots(figsize=(10,6))
# create a plot with a long legend
for ii in range(20):
ax.plot(x, x**2, label='20201110_120000')
ax.plot(x, x, label='20201104_110000')
# grab handles and labels from the first ax and pass it to the second
ax.legend(ncol=2, loc=[1,0])
plt.tight_layout()
Output:

seaborn distplot different bar width on each figure

Sorry for giving an image however I think it is the best way to show my problem.
As you can see all of the bin width are different, from my understanding it shows range of rent_hours. I am not sure why different figure have different bin width even though I didn't set any.
My code looks is as follows:
figure, axes = plt.subplots(nrows=4, ncols=3)
figure.set_size_inches(18,14)
plt.subplots_adjust(hspace=0.5)
for ax, age_g in zip(axes.ravel(), age_cat):
group = total_usage_df.loc[(total_usage_df.age_group == age_g) & (total_usage_df.day_of_week <= 4)]
sns.distplot(group.rent_hour, ax=ax, kde=False)
ax.set(title=age_g)
ax.set_xlim([0, 24])
figure.suptitle("Weekday usage pattern", size=25);
additional question:
Seaborn : How to get the count in y axis for distplot using PairGrid for here it says that kde=False makes y-axis count however http://seaborn.pydata.org/generated/seaborn.distplot.html in the doc, it uses kde=False and still seems to show something else. How can I set y-axis to show count?
I've tried
sns.distplot(group.rent_hour, ax=ax, norm_hist=True) and it still seems to give something else rather than count.
sns.distplot(group.rent_hour, ax=ax, kde=False) gives me count however I don't know why it is giving me count.

Answer 1:
From the documentation:
norm_hist : bool, optional
If True, the histogram height shows a density rather than a count.
This is implied if a KDE or fitted density is plotted.
So you need to take into account your bin width as well, i.e. compute the area under the curve and not just the sum of the bin heights.
Answer 2:
# Plotting hist without kde
ax = sns.distplot(your_data, kde=False)
# Creating another Y axis
second_ax = ax.twinx()
#Plotting kde without hist on the second Y axis
sns.distplot(your_data, ax=second_ax, kde=True, hist=False)
#Removing Y ticks from the second axis
second_ax.set_yticks([])

Subplot charts plotted within a loop very squashed

I have a dataframe with ~120 features that I would like to examine by year. I am plotting each feature, x = year, y = feature value within a loop. Whilst these plot successfully, the charts are illegible as they are totally squashed.
I have tried using plt.tight_layout() and adjusting the figure size using plt.rcParams['figure.figsize'] but sadly to no avail
for i in range(len(roll_df.columns)):
plt.subplot(len(roll_df.columns), 1, i+1)
name = roll_df.columns[i]
plt.plot(roll_df[name])
plt.title(name, y=0)
plt.yticks([])
plt.xticks([])
plt.tight_layout()
plt.show()
The loop runs but all plots are so squashed on the y-axis as to become illegible:

Matplotlib will not automatically adjust the size of your figure. So if you add more subplots below each other, it will split the available space instead of extending the figure. That's why your y axes are so narrow.
You could try to define the figure size beforehand, or determine the figure size based on how many subplots you have:
n_plots = roll_df.shape[1]
fig, axes = plt.subplots(n_plots, 1, figsize=(8, 4 * n_plots), tight_layout=True)
# Then your usual part, but plot on the created axes
for i in range(n_plots):
name = roll_df.columns[i]
axes[i].plot(roll_df[name])
axes[i].title(name, y=0)
axes[i].yticks([])
axes[i].xticks([])
plt.show()

matplotlib: align y-ticks in twinx [duplicate]

I created a matplotlib plot that has 2 y-axes. The y-axes have different scales, but I want the ticks and grid to be aligned. I am pulling the data from excel files, so there is no way to know the max limits beforehand. I have tried the following code.
# creates double-y axis
ax2 = ax1.twinx()
locs = ax1.yaxis.get_ticklocs()
ax2.set_yticks(locs)
The problem now is that the ticks on ax2 do not have labels anymore. Can anyone give me a good way to align ticks with different scales?

Aligning the tick locations of two different scales would mean to give up on the nice automatic tick locator and set the ticks to the same positions on the secondary axes as on the original one.
The idea is to establish a relation between the two axes scales using a function and set the ticks of the second axes at the positions of those of the first.
import matplotlib.pyplot as plt
import matplotlib.ticker
fig, ax = plt.subplots()
# creates double-y axis
ax2 = ax.twinx()
ax.plot(range(5), [1,2,3,4,5])
ax2.plot(range(6), [13,17,14,13,16,12])
ax.grid()
l = ax.get_ylim()
l2 = ax2.get_ylim()
f = lambda x : l2[0]+(x-l[0])/(l[1]-l[0])*(l2[1]-l2[0])
ticks = f(ax.get_yticks())
ax2.yaxis.set_major_locator(matplotlib.ticker.FixedLocator(ticks))
plt.show()
Note that this is a solution for the general case and it might result in totally unreadable labels depeding on the use case. If you happen to have more a priori information on the axes range, better solutions may be possible.
Also see this question for a case where automatic tick locations of the first axes is sacrificed for an easier setting of the secondary axes tick locations.

To anyone who's wondering (and for my future reference), the lambda function f in ImportanceofBeingErnest's answer maps the input left tick to a corresponding right tick through:
RHS tick = Bottom RHS tick + (% of LHS range traversed * RHS range)
Refer to this question on tick formatting to truncate decimal places:
from matplotlib.ticker import FormatStrFormatter
ax2.yaxis.set_major_formatter(FormatStrFormatter('%.2f')) # ax2 is the RHS y-axis

Plotting multiple lines with different scales on the same graph

What is the most pythonic way to plot multiple lineswith very different scales in the same graph with matplotlib? I know can create subplots, but I am not really sure they will give me the best visualization. I don't really care about coloring, legends or any other intricacies at this point.

If you only need two scales then you can simple use twinx and/or twiny
fig, ax = plt.subplots(1, 1)
x = np.arange(11)
ax.plot(x, 'r')
ax2 = ax.twinx()
ax2.plot(x ** 2, 'g')
plt.draw()
I you need more than two see matplotlib: adding second axes() with transparent background? or look into parasitic axes.