Learning Matplot Library¶
Matplot是一个超强的Python绘图库。
在此记录学习Matplot一些例子，以便于日后可以较快的熟悉学习。
在此以量化计算的一些数据为例，如下将EDA计算解离过程的数据保存为一个CSV格式的数据文件，第一列为横坐标，其它列作为纵坐标。
# d, Ees, Eex, Erep, Epol, Edisp, Eh, E
1.3,-47.49,-66.98,135.53,-17.40,-3.69,3.66,-0.03
1.4,-40.61,-48.73,96.73,-8.88,-3.22,-1.50, -4.72
1.5,-35.05,-35.24,68.72,-2.93,-2.76,-4.50, -7.26
1.6,-28.90,-21.83,41.59,2.70,-2.16,-6.43, -8.60
1.8,-24.15,-12.94,24.12,6.26,-1.63,-6.71, -8.34
2.0,-19.85,-6.51,11.83,8.66,-1.12,-5.87, -6.99
2.2,-16.87,-3.23,5.75,9.86,-0.77,-4.49, -5.26
2.5,-13.65,-1.11,1.92,9.95,-0.44,-2.88, -3.33
2.8,-11.36,-0.38,0.64,9.25,-0.26,-1.85, -2.11
3.1,-9.70,-0.13,0.21,8.56,-0.16,-1.06, -1.22
3.5,-8.07,-0.03,0.05,7.63,-0.09,-0.42, -0.51
4.0,-6.61,0.00,0.01,6.52,-0.05,-0.09, -0.13
4.5,-5.45,0.00,0.00,5.38,-0.02,-0.07, -0.09
5.0,-4.56,0.00,0.00,4.51,-0.01,-0.05, -0.07
6.0,-3.34,0.00,0.00,3.30,0.00,-0.04, -0.04
7.0,-2.56,0.00,0.00,2.53,0.00,-0.03, -0.03
8.0,-2.02,0.00,0.00,2.00,0.00,-0.02, -0.02
9.0,-1.64,0.00,0.00,1.62,0.00,-0.02, -0.02
10.0,-1.35,0.00,0.00,1.34,0.00,-0.02, -0.01
用于绘图的python代码如下,此部分的重点在如何拟合光滑的曲线。
直接法¶
这部分我将绘图的相关信息直接写入在python代码中。
#!/bin/env python2
# -*- encoding:utf-8 -*-
import numpy as np
import matplotlib.pyplot as plt
from scipy.interpolate import spline
from scipy import interpolate
plt.figure(figsize=(8,4), dpi=300)
#plt.figure(dpi=300)
dataFile = 'summary'
data = np.genfromtxt(dataFile, delimiter=',').transpose()
newx = np.linspace(data[0].min(), data[0].max(), 500)
#for kind in ['nearest', 'zero', 'slinear', 'quadratic', 'cubic']:
for kind in ['cubic']:
f = interpolate.interp1d(data[0], data[7], kind=kind)
newy = f(newx)
plt.plot(newx, newy, label=kind)
plt.plot(data[0], data[7], '+')
plt.title('The Interaction Energy Between $NH_4$...$OH_2$')
plt.xlabel('d/A')
plt.ylabel('E/$kcal.mol^{-1}$')
plt.xlim([1, 10.5])
plt.ylim([-9, 0.2])
plt.legend(loc='lower right')
plt.savefig('test1.png', dpi=120)
对其中的函数进行简单的简介：
- numpy.genfromtxt 从文件中读取数据
- numpy.linspace 在某一数据段产生一系列均匀的数据。用于后面拟合滑曲线。
- scipy.interpolate.interp1d 使用指定算法对数据进行拟合，得到一个函数。
- pyplot.figure 用于设定画布相关属性
- matplotlib.pyplot.plot 绘制图形
- matplotlib.pyplot.title 设定图形的标题
- matplotlib.pyplot.xlabel, ylabel 分别设定横坐标、纵坐标的标签
- matplotlib.pyplot.xlim, ylim 分别设定横坐标、纵坐标的取值范围
- matplotlib.pyplot.legend 设定关于图例的一些属性
- matplotlib.pyplot.savefig 输出图形至文件
使用json作配置文件¶
这里我将绘图的一些配置信息以配置文件的方式存放，绘制图形时从配置文件中读取数据。最初愿望是实现数据，程序，配置三者分离，便于通用的处理大量数据。配置文件为json格式，如下：
{
"inputfile": {
"name":"summary",
"delimiter":","
},
"outputfile": {
"name":"test2.png",
"dpi":120
},
"title": "The Interaction Energy Between $NH_4$...$OH_2$",
"xlabel": "d/A",
"ylabel": "E/$kcal.mol^{-1}$",
"xrange": [1, 10.5],
"yrange": [-9, 0.2],
"legend": {
"location": "lower right"
}
}
绘图代码略有修改，如下:
#!/bin/env python2
# -*- encoding:utf-8 -*-
#
# Author: Liu Hui
# Date: Thu Mar 14 23:21:42 CST 2013
#
# 利用matplotlib图形库来处理数据作图。需要安装matplotlib包。
#
# 此处将作图的一些配置文件写为json的格式导入，希望可以做到数据，程序，控制分离
#
import json
import numpy as np
import matplotlib.pyplot as plt
from scipy.interpolate import spline
from scipy import interpolate
f = open('config.json','r')
config = json.load(f)
f.close()
data = np.genfromtxt(config['inputfile']['name'],
delimiter=config['inputfile']['delimiter']).transpose()
newx = np.linspace(data[0].min(), data[0].max(), 500)
#for kind in ['nearest', 'zero', 'slinear', 'quadratic', 'cubic']:
for kind in ['cubic']:
f = interpolate.interp1d(data[0], data[7], kind=kind)
newy = f(newx)
plt.plot(newx, newy, label=kind)
plt.plot(data[0], data[7], '+')
plt.title(config['title'])
plt.xlabel(config['xlabel'])
plt.ylabel(config['ylabel'])
plt.xlim(config['xrange'])
plt.ylim(config['yrange'])
plt.legend(loc=config['legend']['location'])
plt.savefig(config['outputfile']['name'], dpi=config['outputfile']['dpi'])