Examples of plotannot annotations

This notebooks presents a few examples of the plotannot package for annotating axis labels.

Getting ready

#Load a few packages for plotting
import numpy as np
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
np.random.seed(1) #make notebook reproducible

#Load plotannot
import plotannot

#Create some fake data
table = pd.DataFrame(np.random.random((100,50)))
table.head()

	0	1	2	3	4	5	6	7	8	9	...	40	41	42	43	44	45	46	47	48	49
0	0.417022	0.720324	0.000114	0.302333	0.146756	0.092339	0.186260	0.345561	0.396767	0.538817	...	0.988861	0.748166	0.280444	0.789279	0.103226	0.447894	0.908596	0.293614	0.287775	0.130029
1	0.019367	0.678836	0.211628	0.265547	0.491573	0.053363	0.574118	0.146729	0.589306	0.699758	...	0.114746	0.949489	0.449912	0.578390	0.408137	0.237027	0.903380	0.573679	0.002870	0.617145
2	0.326645	0.527058	0.885942	0.357270	0.908535	0.623360	0.015821	0.929437	0.690897	0.997323	...	0.556240	0.136455	0.059918	0.121343	0.044552	0.107494	0.225709	0.712989	0.559717	0.012556
3	0.071974	0.967276	0.568100	0.203293	0.252326	0.743826	0.195429	0.581359	0.970020	0.846829	...	0.263297	0.065961	0.735066	0.772178	0.907816	0.931972	0.013952	0.234362	0.616778	0.949016
4	0.950176	0.556653	0.915606	0.641566	0.390008	0.485991	0.604310	0.549548	0.926181	0.918733	...	0.315245	0.892889	0.577857	0.184010	0.787929	0.612031	0.053909	0.420194	0.679069	0.918602

5 rows × 50 columns

#Prepare some english words to use for labels
import english_words
words = sorted(list(english_words.english_words_lower_set))
rand_integers = np.random.randint(0, len(words), size=100)
words = [words[i] for i in rand_integers] #100 random words

Simple example

#Plot heatmap
ax = sns.heatmap(table, xticklabels=True, yticklabels=False)

#Rotate all labels
plotannot.format_ticklabels(ax, axis="xaxis", rotation=45)

#Annotate labels
to_label = range(20,35)
plotannot.annotate_ticks(ax, axis="xaxis", labels=to_label)

#Color individual labels
plotannot.format_ticklabels(ax, axis="xaxis", labels=[25], color="red")

plt.savefig("simple_example.png", bbox_inches="tight", facecolor='white')

Annotate axis of scatter plots

fig, ax = plt.subplots()
ax.scatter(words, range(len(words)))

#First rotate to make labels fit
plotannot.format_ticklabels(ax, axis="xaxis", rotation=90)

#Annotate labels
to_label = [words[i] for i in range(20,35)]
plotannot.annotate_ticks(ax, axis="xaxis", labels=to_label, rel_label_size=2) #rel_label_size gives a bit better space for the labels

#Finally, rotate labels to be more visible
plotannot.format_ticklabels(ax, axis="xaxis", rotation=30, ha="right")

Annotate x- and/or y-axis ticks

#Plot data with labels
fig, (ax1, ax2) = plt.subplots(1,2, figsize=(10,5))
sns.heatmap(table, yticklabels=True, xticklabels=True, cbar=False, ax=ax1)
sns.heatmap(table, yticklabels=True, xticklabels=True, cbar=False, ax=ax2)

#Apply annotation to second axes
to_label = range(10,20)
plotannot.annotate_ticks(ax2, axis="xaxis", labels=to_label)
plotannot.annotate_ticks(ax2, axis="yaxis", labels=to_label)

#Label axes
_ = ax1.set_title("Before annotation", size=15)
_ = ax2.set_title("After annotation", size=15)

plt.savefig("before_after.png", bbox_inches="tight", facecolor='white')

The function annotate_ticks also works directly on seaborn objects such as the ClusterGrid of sns.clustermap:

g = sns.clustermap(table, yticklabels=True, xticklabels=True, figsize=(4,4))

to_label = range(10,20)
plotannot.annotate_ticks(g, axis="xaxis", labels=to_label)
plotannot.annotate_ticks(g, axis="yaxis", labels=to_label)

Choose specific labels for bottom/top and left/right axes

#Plot data and turn on all labels on
ax = sns.heatmap(table, yticklabels=True, xticklabels=True, cbar=False)
ax.tick_params(right=True, top=True, left=True, bottom=True,
               labelleft=True, labelbottom=True, labelright=True, labeltop=True)

#Apply annotation for each axis
left_labels = range(10)
plotannot.annotate_ticks(ax, axis="left", labels=left_labels)

right_labels = range(90,100)
plotannot.annotate_ticks(ax, axis="right", labels=right_labels)

bottom_labels = range(10,16)
plotannot.annotate_ticks(ax, axis="bottom", labels=bottom_labels)

top_labels = range(23,30)
plotannot.annotate_ticks(ax, axis="top", labels=top_labels)

Highlight specific rows with color, size and annotation

ax = sns.heatmap(table, xticklabels=False, yticklabels=True, cbar=False)

#Color a subset of labels
to_format = [2,3,5,7,11,13,17,19,23,29]
plotannot.format_ticklabels(ax, "yaxis", labels=to_format, color="red", fontsize=20)

#Annotate all ticks
plotannot.annotate_ticks(ax, "yaxis", labels=range(0,30), expand_axis=1, perp_shift=10)

Highlight and rotate labels

#Plot data
word_table = pd.DataFrame(np.random.random((50,100)), columns=words)
ax = sns.heatmap(word_table, yticklabels=False, xticklabels=True)

#Label specific words
plotannot.format_ticklabels(ax, "xaxis", labels=["mail"], color="red", format_ticks=True, fontsize=16)
plotannot.format_ticklabels(ax, "xaxis", labels=["methyl"], color="blue", format_ticks=True, fontsize=16)

#Annotate ticks starting with "m"
m_words = [word for word in words if word.startswith("m")]
plotannot.annotate_ticks(ax, "xaxis", labels=m_words, rel_label_size=1.5)

#Rotate all ticklabels to 45 degrees
plotannot.format_ticklabels(ax, "xaxis", rotation=45, ha="right")