objectif

Carte de chaleur de la matrice de corrélation.

code minimal

import seaborn as sns, pandas as pd, numpy as np
df = pd.DataFrame(np.random.default_rng(0).normal(size=(20,4)), columns=list("abcd"))
ax = sns.heatmap(df.corr())
print(hasattr(ax, "figure"))

utilisation

import seaborn as sns, pandas as pd, numpy as np
df = pd.DataFrame(np.random.default_rng(1).normal(size=(10,3)), columns=list("xyz"))
ax = sns.heatmap(df.corr(), annot=False)
print("ok")

variante(s) utile(s)

import seaborn as sns, pandas as pd, numpy as np
ax = sns.heatmap(pd.DataFrame([[1,0],[0,1]]))
print("ok")

notes

• Standardiser si échelles hétérogènes avant corr.