Annotated heatmap Example

Here shows how to draw annotated heatmap

import numpy as np
import pandas as pd
import milkviz as mv

np.random.seed(0)

First let’s create some random data

import seaborn as sns

p = sns.load_dataset("penguins").dropna()
df = p.set_index(["species", "island", "sex"])
data = np.random.randn(*df.to_numpy().shape)
df = pd.DataFrame(data=data, columns=df.columns, index=df.index)
df.columns.name = "col"
df

		col	bill_length_mm	bill_depth_mm	flipper_length_mm	body_mass_g
species	island	sex
Adelie	Torgersen	Male	1.764052	0.400157	0.978738	2.240893
		Female	1.867558	-0.977278	0.950088	-0.151357
		Female	-0.103219	0.410599	0.144044	1.454274
		Female	0.761038	0.121675	0.443863	0.333674
		Male	1.494079	-0.205158	0.313068	-0.854096
...	...	...	...	...	...	...
Gentoo	Biscoe	Female	-0.255958	-0.348046	-0.782367	0.625119
		Female	-0.813596	-0.521642	-0.073120	-1.297380
		Male	-0.324935	-0.711306	-0.388154	-0.059928
		Female	-0.799914	-0.220076	1.308669	-0.025799
		Male	1.145262	0.346494	0.774161	-0.774459

333 rows × 4 columns

Create the heatmap

mv.anno_clustermap(df, row_colors=["species", "island", "sex"],
                   col_label="col", z_score=0)

<seaborn.matrix.ClusterGrid object at 0x7f2cc233aca0>