简单散点密度图

使用hist2d绘制2维直方图

from matplotlib.colors import LogNorm
import matplotlib.pyplot as plt
import numpy as np

# normal distribution center at x=0 and y=5
x = np.random.randn(100000)
y = np.random.randn(100000) + 5

plt.hist2d(x, y, bins=40, norm=LogNorm())
plt.colorbar()
plt.show()

边缘箱形图 （Marginal Boxplot）

散点图加右侧和底部直方图。功能实现是先形成4X4的网络底图，使用左上3X3画散点图，右侧3X1和底部1X3画直方图。

# !pip install brewer2mpl
import numpy as np
import pandas as pd
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns
import warnings; warnings.filterwarnings(action='once')
import os

os.environ['NLS_LANG'] = 'AMERICAN_AMERICA.ZHS16GBK'

config = {'font.family': 'serif',
            'font.serif': 'FangSong',
            'font.style': 'italic',
            #'font.weight': 'blod',#'normal',  # or 'blod'
            'font.size': 10,  # 此处貌似不能用类似large、small、medium字符串
            'axes.unicode_minus': False
            }
plt.rcParams.update(config)
#plt.rcParams['font.sans-serif']=['SimHei']
plt.rcParams['axes.unicode_minus']= False
plt.style.use('seaborn-whitegrid')
sns.set_style("white")
# Import Data
df = pd.read_csv("./mpg_ggplot2.csv")
import matplotlib as mpl
mpl.rcParams['font.sans-serif'] = ['SimHei']
mpl.rcParams['font.serif'] = ['SimHei']
# Create Fig and gridspec
fig = plt.figure(figsize=(16, 10), dpi= 200)
grid = plt.GridSpec(4, 4, hspace=0.5, wspace=0.2)

# Define the axes
ax_main = fig.add_subplot(grid[:-1, :-1])
ax_right = fig.add_subplot(grid[:-1, -1], xticklabels=[], yticklabels=[])
ax_bottom = fig.add_subplot(grid[-1, 0:-1], xticklabels=[], yticklabels=[])

# Scatterplot on main ax
ax_main.scatter('displ', 'hwy', s=df.cty*4, c=df.manufacturer.astype('category').cat.codes, alpha=.9, data=df, cmap="tab10", edgecolors='gray', linewidths=.5)

# histogram on the right
ax_bottom.hist(df.displ, 40, histtype='stepfilled', orientation='vertical', color='deeppink')
ax_bottom.invert_yaxis()

# histogram in the bottom
ax_right.hist(df.hwy, 40, histtype='stepfilled', orientation='horizontal', color='deeppink')

# Decorations
ax_main.set(title='边缘箱形图 \n displ vs hwy', xlabel='displ', ylabel='hwy')
ax_main.title.set_fontsize(20)
for item in ([ax_main.xaxis.label, ax_main.yaxis.label] + ax_main.get_xticklabels() + ax_main.get_yticklabels()):
    item.set_fontsize(14)

#xlabels = ax_main.get_xticks().tolist()
#ax_main.set_xticklabels(xlabels)
plt.savefig('Scatterplot_with_Histograms.png',dpi=200)

散点密度图加直方图

使用scipy.stats.gaussian_kde进行核密度进行计算，这是使用Kernel density estimation的方法，估计出每个位置的密度值。 这代码参考森气笔记的代码，原代码有点问题，出图时显示不正确（下面图1），右侧和底部直方统计图位置不正确，根据上面4X4网络底图方法，稍做修改，得到正确的图。

# -*- encoding: utf-8 -*-
import pandas as pd
import os
import matplotlib.pyplot as plt
from scipy.stats import gaussian_kde
from mpl_toolkits.axes_grid1 import make_axes_locatable
import numpy as np
from matplotlib import ticker, cm
import seaborn as sns

df = pd.read_csv('./Fig1a_XY_obs.csv',index_col=0)
x,y = df.CxP.values,df.FR.values
xy = np.vstack([x,y])#将两个维度的数据进行叠加
kenal = gaussian_kde(xy)#这一步根据xy这个样本数据，在全定义域上建立了概率密度分布，所以kenal其实就是一个概率密度函数，输入对应的(x,y)坐标，就给出相应的概率密度
#kenal.set_bandwidth(bw_method=kenal1.factor)#这一步可以直接设置bandwith，通常情况下默认即可
z = kenal.evaluate(xy)#得到我们每个样本点的概率密度
z = gaussian_kde(xy)(xy)#这行代码和上面两行是相同的意思，这行是一行的写法
idx = z.argsort() #对z值进行从小到大排序并返回索引
x, y, z = x[idx], y[idx], z[idx]#对x,y按照z的升序进行排列
#上面两行代码是为了使得z值越高的点，画在上面，不被那些z值低的点挡住，从美观的角度来说还是十分必要的
#fig, ax = plt.subplots(figsize=(6,6),dpi=100)
fig = plt.figure(figsize=(8, 8), dpi= 200)
grid = plt.GridSpec(4, 4, hspace=0.5, wspace=0.2)
ax = fig.add_subplot(grid[:-1, :-1])
ax_histy = fig.add_subplot(grid[:-1, -1], xticklabels=[], yticklabels=[])
ax_histx = fig.add_subplot(grid[-1, 0:-1], xticklabels=[], yticklabels=[])
xcord=x
ycord=y
m = ((xcord*ycord).mean() - xcord.mean()* ycord.mean())/(pow(xcord,2).mean()-pow(xcord.mean(),2))
c = ycord.mean() - m*xcord.mean()
scatter=ax.scatter(x,y,c=z,s = 2)
divider=make_axes_locatable(ax)
ax.set_xlim(0,0.01)
ax.set_ylim(0,1.5)
ax.set_yticks(np.arange(0,1.51,0.5))
#ax_histx = divider.append_axes("bottom", 1.2, pad=0.15)
#ax_histy = divider.append_axes("right", 1.2, pad=0.15)
ax.set_xticklabels([])
# line 
ds = pd.read_csv('./Fig1a_XY_mod.csv',index_col=0)

ys = ['Y_pl', 'Y_pl_op', 'Y_sc', 'Y_li1', 'Y_np', 'Y_mean']
linestyles = ['-.','-.','-.','-.','-.','-']
colors = ['#723BA0','#8575D7','#8DCAD0','#E0C671','#9F795D','k']
labels = ['Power law','Power law (linear opt.)','Scale','Linear','Non-parametric','Mean']
for i in range(len(ys)):
    y = ds[ys[i]].values
    ax.plot(ds.X,y,linestyle=linestyles[i],c = colors[i],label = labels[i])
    ax.legend(frameon=False)
ax.set_ylabel('Lightning flash rate (number per km$^{2}$ per month)')
ax.set_xticks([0,0.005,0.010])

# part b
df2 = pd.read_csv('./Fig1b_CxP.csv',index_col=0)
sns.histplot(df2['rcp'].values, kde=True,ax = ax_histx,edgecolor = 'w',bins=51)
sns.histplot(df2['hist'].values, kde=True,ax = ax_histx,edgecolor = 'w',bins=30,color = 'red')

ax_histx.plot(1,1,label = 'Present day',c = '#D4A663')
ax_histx.plot(1,1,label = 'Future',c = '#0202FF')
ax_histx.legend(frameon=False)
ax_histx.set_ylabel('Norm. dist.')
ax_histx.set_xlabel('CAPE × Precip (W m$^{2}$)')

ax_histx.set_xticks([0,0.005,0.010])
ax_histx.set_xticklabels([0,0.005,'0.010'])
ax_histx.set_xlim(0,0.01)
ax_histx.set_ylim(0,400)
ax_histx.set_yticks([0,200,400])
# part c
df3 = pd.read_csv('./Fig1c_FR.csv',index_col=0)
ax_histy.set_xlabel('Norm. dist.')
ax_histy.set_yticks([])
ax_histy = ax_histy.twinx()
ax_histy.set_ylabel('Lightning flash rate (number per km$^{2}$ per month)',rotation = 270,labelpad = 15)
sns.histplot(y = df3['rcp'].values, kde=True,ax = ax_histy,edgecolor = 'w',bins=30)
sns.histplot(y = df3['hist'].values, kde=True,ax = ax_histy,color = 'red',edgecolor = 'w',bins=30)
ax_histy.set_xlim(350,0)
ax_histy.set_xticklabels([0,2.5])
ax_histy.set_ylim(0,1.5)
ax_histy.set_yticks(np.arange(0,1.51,0.5))
ax_histy.set_xlabel('Norm. dist.')
plt.rcParams['xtick.bottom'] = plt.rcParams['xtick.labelbottom'] = True
plt.rcParams['xtick.top'] = plt.rcParams['xtick.labeltop'] = False
plt.savefig('png2.png',dpi = 600)
#plt.show()