Python绘制散点密度图的三种方式详解
作者:气象水文科研猫
散点密度图是在散点图的基础上,计算了每个散点周围分布了多少其他的点,并通过颜色表现出来。本文主要介绍了Python绘制散点密度图的三种方式,需要的可以参考下
方式一
import matplotlib.pyplot as plt import numpy as np from scipy.stats import gaussian_kde from mpl_toolkits.axes_grid1 import make_axes_locatable from matplotlib import rcParams config = {"font.family":'Times New Roman',"font.size": 16,"mathtext.fontset":'stix'} rcParams.update(config) # 读取数据 import pandas as pd filename=r'F:/Rpython/lp37/testdata.xlsx' df2=pd.read_excel(filename)#读取文件 x=df2['data1'].values y=df2['data2'].values xy = np.vstack([x,y]) z = gaussian_kde(xy)(xy) idx = z.argsort() x, y, z = x[idx], y[idx], z[idx] fig,ax=plt.subplots(figsize=(12,9),dpi=100) scatter=ax.scatter(x,y,marker='o',c=z,edgecolors='',s=15,label='LST',cmap='Spectral_r') cbar=plt.colorbar(scatter,shrink=1,orientation='vertical',extend='both',pad=0.015,aspect=30,label='frequency') #orientation='horizontal' font3={'family':'SimHei','size':16,'color':'k'} plt.ylabel("估计值",fontdict=font3) plt.xlabel("预测值",fontdict=font3) plt.savefig('F:/Rpython/lp37/plot70.png',dpi=800,bbox_inches='tight',pad_inches=0) plt.show()
方式二
from statistics import mean import matplotlib.pyplot as plt from sklearn.metrics import explained_variance_score,r2_score,median_absolute_error,mean_squared_error,mean_absolute_error from scipy import stats import numpy as np from matplotlib import rcParams config = {"font.family":'Times New Roman',"font.size": 16,"mathtext.fontset":'stix'} rcParams.update(config) def scatter_out_1(x,y): ## x,y为两个需要做对比分析的两个量。 # ==========计算评价指标========== BIAS = mean(x - y) MSE = mean_squared_error(x, y) RMSE = np.power(MSE, 0.5) R2 = r2_score(x, y) MAE = mean_absolute_error(x, y) EV = explained_variance_score(x, y) print('==========算法评价指标==========') print('BIAS:', '%.3f' % (BIAS)) print('Explained Variance(EV):', '%.3f' % (EV)) print('Mean Absolute Error(MAE):', '%.3f' % (MAE)) print('Mean squared error(MSE):', '%.3f' % (MSE)) print('Root Mean Squard Error(RMSE):', '%.3f' % (RMSE)) print('R_squared:', '%.3f' % (R2)) # ===========Calculate the point density========== xy = np.vstack([x, y]) z = stats.gaussian_kde(xy)(xy) # ===========Sort the points by density, so that the densest points are plotted last=========== idx = z.argsort() x, y, z = x[idx], y[idx], z[idx] def best_fit_slope_and_intercept(xs, ys): m = (((mean(xs) * mean(ys)) - mean(xs * ys)) / ((mean(xs) * mean(xs)) - mean(xs * xs))) b = mean(ys) - m * mean(xs) return m, b m, b = best_fit_slope_and_intercept(x, y) regression_line = [] for a in x: regression_line.append((m * a) + b) fig,ax=plt.subplots(figsize=(12,9),dpi=600) scatter=ax.scatter(x,y,marker='o',c=z*100,edgecolors='',s=15,label='LST',cmap='Spectral_r') cbar=plt.colorbar(scatter,shrink=1,orientation='vertical',extend='both',pad=0.015,aspect=30,label='frequency') plt.plot([0,25],[0,25],'black',lw=1.5) # 画的1:1线,线的颜色为black,线宽为0.8 plt.plot(x,regression_line,'red',lw=1.5) # 预测与实测数据之间的回归线 plt.axis([0,25,0,25]) # 设置线的范围 plt.xlabel('OBS',family = 'Times New Roman') plt.ylabel('PRE',family = 'Times New Roman') plt.xticks(fontproperties='Times New Roman') plt.yticks(fontproperties='Times New Roman') plt.text(1,24, '$N=%.f$' % len(y), family = 'Times New Roman') # text的位置需要根据x,y的大小范围进行调整。 plt.text(1,23, '$R^2=%.3f$' % R2, family = 'Times New Roman') plt.text(1,22, '$BIAS=%.4f$' % BIAS, family = 'Times New Roman') plt.text(1,21, '$RMSE=%.3f$' % RMSE, family = 'Times New Roman') plt.xlim(0,25) # 设置x坐标轴的显示范围 plt.ylim(0,25) # 设置y坐标轴的显示范围 plt.savefig('F:/Rpython/lp37/plot71.png',dpi=800,bbox_inches='tight',pad_inches=0) plt.show()
方式三
import pandas as pd import numpy as np from scipy import optimize import matplotlib.pyplot as plt from matplotlib import cm from matplotlib.colors import Normalize from scipy.stats import gaussian_kde from matplotlib import rcParams config={"font.family":'Times New Roman',"font.size":16,"mathtext.fontset":'stix'} rcParams.update(config) # 读取数据 filename=r'F:/Rpython/lp37/testdata.xlsx' df2=pd.read_excel(filename)#读取文件 x=df2['data1'].values.ravel() y=df2['data2'].values.ravel() N = len(df2['data1']) #绘制拟合线 x2 = np.linspace(-10,30) y2 = x2 def f_1(x,A,B): return A*x + B A1,B1 = optimize.curve_fit(f_1,x,y)[0] y3 = A1*x + B1 # Calculate the point density xy = np.vstack([x,y]) z = gaussian_kde(xy)(xy) norm = Normalize(vmin = np.min(z), vmax = np.max(z)) #开始绘图 fig,ax=plt.subplots(figsize=(12,9),dpi=600) scatter=ax.scatter(x,y,marker='o',c=z*100,edgecolors='',s=15,label='LST',cmap='Spectral_r') cbar=plt.colorbar(scatter,shrink=1,orientation='vertical',extend='both',pad=0.015,aspect=30,label='frequency') cbar.ax.locator_params(nbins=8) cbar.ax.set_yticklabels([0.005,0.010,0.015,0.020,0.025,0.030,0.035])#0,0.005,0.010,0.015,0.020,0.025,0.030,0.035 ax.plot(x2,y2,color='k',linewidth=1.5,linestyle='--') ax.plot(x,y3,color='r',linewidth=2,linestyle='-') fontdict1 = {"size":16,"color":"k",'family':'Times New Roman'} ax.set_xlabel("PRE",fontdict=fontdict1) ax.set_ylabel("OBS",fontdict=fontdict1) # ax.grid(True) ax.set_xlim((0,25)) ax.set_ylim((0,25)) ax.set_xticks(np.arange(0,25.1,step=5)) ax.set_yticks(np.arange(0,25.1,step=5)) plt.savefig('F:/Rpython/lp37/plot72.png',dpi=800,bbox_inches='tight',pad_inches=0) plt.show()
到此这篇关于Python绘制散点密度图的三种方式详解的文章就介绍到这了,更多相关Python散点密度图内容请搜索脚本之家以前的文章或继续浏览下面的相关文章希望大家以后多多支持脚本之家!