超前滞后相关关系的热力图

功能:

1. 热力图常用于展示多个气象变量之间的两两相关性（横纵坐标均为变量名称），也可用于展示两个气象变量之间的超前滞后相关关系（纵坐标为变量1的时间，横坐标为变量2超前于变量1的时间）。示例图为ACE和SIC在7-9月的十天尺度的超前滞后相关关系

程序清单：

#导入包

import numpy as np

import pandas as pd

import xarray as xr

import seaborn as sns

import matplotlib.pyplot as plt

from matplotlib import colors

#读入数据

data = xr.open_dataset('F:/python/data/6.21.7_corr.nc')

corr = data['corr']

p = data['p']

#绘图

#统一修改字体为Arial

plt.rcParams['font.family'] = ['Arial']    

#定义绘制热力图的函数

def heatmap(data,ax):

    len_x = int(len(data[0,:])/2)

    sns.heatmap(data,

          vmin = -0.6,  #colobar最小值

          vmax = 0.6,  #colobar最大值

          cmap = colors.ListedColormap(pd.read_csv('BlueDarkRed182.rgb',sep='\s+').values/255),  #颜色

          fmt = '.2f',  #值保留两位小数

          annot = True,

          annot_kws = {'size':7,'weight':'bold'},  #定义方框内的文字大小，设置加粗

          linewidths = 0,    #设置方框边缘线宽度

          linecolor = 'white',    #设置方框边缘线颜色

          cbar = True,    #显示colorbar

          cbar_kws={'pad': 0.02,'shrink': 0.9},    #设置colorbar格式，pad为距主图的距离，shrink为缩放比例

          cbar_ax = None,

          square = True,    #设置方框为正方形

          xticklabels = np.arange(-len_x,len_x+1)*10,    #设置横坐标

          yticklabels =[ '7.0$\mathregular{1^{st}}$–7.1$\mathregular{0^{th}}$',

                         '7.1$\mathregular{1^{th}}$–7.2$\mathregular{0^{th}}$',

                         '7.2$\mathregular{1^{st}}$–7.3$\mathregular{1^{st}}$',

                         '8.0$\mathregular{1^{st}}$–8.1$\mathregular{0^{th}}$',

                         '8.1$\mathregular{1^{th}}$–8.2$\mathregular{0^{th}}$',

                         '8.2$\mathregular{1^{st}}$–8.3$\mathregular{1^{st}}$',

                         '9.0$\mathregular{1^{st}}$–9.1$\mathregular{0^{th}}$',

                         '9.1$\mathregular{1^{th}}$–9.2$\mathregular{0^{th}}$',

                         '9.2$\mathregular{1^{st}}$–9.3$\mathregular{0^{th}}$',],#设置纵坐标

          mask = None,

          ax = ax,

    )

    for label in ax.get_yticklabels():

        label.set_rotation(360)  #y轴坐标旋转

    return ax

fig = plt.figure(figsize=(13, 5),dpi=300)

ax=fig.add_subplot(1,1,1)

#调用热力图函数

heatmap(corr,ax)

#添加色标

cbar = ax.collections[0].colorbar

#显著性检验可视化：在置信度水平为99%的值上方添加***，95-99%的值上方添加**，90-95%的值上方添加*

widthx,widthy = 0, -0.12

for n in ax.get_xticks():

    for m in ax.get_yticks():

        pv = (p[int(m),int(n)])

        if  pv< 0.1 and pv>= 0.05:

            ax.text(n+widthx,m+widthy,'*',ha = 'center',color = 'w',size=8, weight='bold')

        if  pv< 0.05 and pv>= 0.01:

            ax.text(n+widthx,m+widthy,'**',ha = 'center',color = 'w',size=8, weight='bold')

        if  pv< 0.01:

            ax.text(n+widthx,m+widthy,'***',ha = 'center',color = 'w',size=8, weight='bold')

#添加横纵坐标

ax.set_xlabel('SIC Lead Time (Unit: day)',loc='center',fontsize=12)   

ax.set_ylabel('ACE Periods',loc='center',fontsize=12)                    

plt.show()