seaborn heatmap

%matplotlib inline

import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns

# f1: 默认
data = np.random.rand(4,4);print(data)
sns.heatmap(data);
# output
[[0.66551277 0.53291346 0.58329178 0.03840753]
 [0.28192606 0.35637815 0.66603227 0.63488134]
 [0.98989239 0.63482035 0.76583562 0.45823605]
 [0.40120982 0.7955449  0.74902682 0.28705992]]

# f2: 设置颜色版的取值范围
sns.heatmap(data, vmin=0.2, vmax=0.5);

# f3: 以0为分界，展示正负数据
data = np.random.randn(4,4);print(data)
sns.heatmap(data, center=0);

# 航班数据集
flight = sns.load_dataset("flights")
flight.head(3)
year	month	passengers
0	1949	January	 112
1	1949	February 118
2	1949	March	 132

# 数据透视表
flight = flight.pivot(index='month', columns='year', values='passengers')
flight.head(3)
year	1949	1950	1951	1952	1953	1954	1955	1956	1957	1958	1959	1960
month												
January	112	115	145	171	196	204	242	284	315	340	360	417
February	118	126	150	180	196	188	233	277	301	318	342	391
March	132	141	178	193	236	235	267	317	356	362	406	419

# f1: default
ax = sns.heatmap(flight)
# f2: 格子中显示数值
ax = sns.heatmap(flight, annot=True, fmt="d")
# f3: 增加空隙
ax = sns.heatmap(flight, linewidth=.2)
# f4: 设置调色板
ax = sns.heatmap(flight, cmap="YlGnBu")