pandas 绘制组合图随手笔记

虽然使用matplotlib可以画出各种组合图形，但是使用pandas自带matplotlib API能更简单高效的绘制出好看的图形。

简单图形绘制

使用pandas自带的api绘制图形非常方便。
导入数据：

In [4]: import pandas as pd

In [5]: import numpy as np

In [6]: import matplotlib.pyplot as plt

In [7]: df = pd.read_excel('01_test_group_stat.xlsx',header=0)

In [8]: df.head()
Out[8]:
   NUM   90dB_std  90dB_mean  90dB_interval   85dB_std  85dB_mean  85dB_interval   80dB_std  80dB_mean      ...        15dB_std  15dB_mean  15dB_interval   10dB_std  10dB_mean  10dB_interval    0dB_std   0dB_mean  0dB_interval
0  101  25.444129 -10.295017         135.04  15.075547   4.539758          56.08  27.982940 -36.032642      ...       30.737979 -11.378675         167.09   8.186763  -9.021683          43.87   7.201149  -1.024150         27.46
1  102   7.111787  17.771917          33.57   7.704128  19.269800          35.48  10.148673  15.964733      ...        9.069287  34.272467          39.67   9.329397 -14.305033          41.96  13.663520  -9.728942         52.26
2  103   4.667749   1.017975          27.08  28.544152  20.333050         135.81   3.415473  25.697400      ...        4.400247   4.619142          26.32  34.558291   0.940050         170.13  10.685239   5.533150         51.11
3  104  14.674394  -2.259317          49.97  13.053583  26.952875          53.40  25.713856  11.828242      ...       26.840332 -32.337458         150.30   8.092768 -20.856675          44.25   4.520228  39.771008         40.05
4  105   4.015146   9.256283          22.51  13.710575   0.719875          54.93   9.083101  32.437342      ...        9.681165 -19.592317          40.44   9.873788  27.430317          42.72  10.993832  33.901133         48.83

[5 rows x 58 columns]

画一些简单的图：

# 密度图
In [10]: df.iloc[:,1].plot(kind='density',legend=True,title='hello world!')
Out[10]: <matplotlib.axes._subplots.AxesSubplot at 0x1dad1a5ed68>

In [11]: plt.xlabel('hello hello')
Out[11]: Text(0.5,0,'hello hello')

In [12]: plt.show()

image.png

#柱状图
In [13]: df.iloc[:,1].plot(kind='hist',legend=True,title='hello world!')
Out[13]: <matplotlib.axes._subplots.AxesSubplot at 0x1dad213d208>

In [14]: plt.show()

image.png

#散点图
In [17]: df.plot(kind='scatter',x='90dB_std',y='90dB_mean',legend=True,title='hello world!')
Out[17]: <matplotlib.axes._subplots.AxesSubplot at 0x1dad49d7208>

In [18]: plt.show()

绘制稍微复杂一些的组合图形

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np

plt.style.use('ggplot')
font = {'family': 'serif',
        'color':  'darkred',
        'weight': 'normal',
        'size': 16,
        }
df_1 = pd.read_csv('./01/tmp/01_90dB_mean_norm_w10_s5.csv', header=0)
df_2 = pd.read_csv('./01/01_stimulate_mean_normed.csv', header=0)
# 当组合图只有一行或者一列时，默认会将axes压成一维的,squeeze参数可取消压缩维度
fig, axes = plt.subplots(ncols=1, nrows=1, figsize=(15, 10), squeeze=False)

for col in list(df_1.columns):
    # label参数表示图例名称
    axes[0, 0].plot(df_1[col], label=col)
  
axes[0, 0].plot(df_2['90dB_mean'], label='90dB_mean')
# 绘制整个fig的图例，而非某个subplot的图例
handles, labels = axes[0, 0].get_legend_handles_labels()
fig.legend(handles, labels)
plt.show()

将x轴的数字序列改为字符串标签

import matplotlib.pyplot as plt
plt.style.use('ggplot') 
font = {'family': 'serif', 'color': 'black', 'weight': 'normal', 'size': 14}

axes = plt.figure().add_subplot(111)
plt.bar(x=[1,2,3], height=[4,5,6])
plt.xticks([1,2,3])
x_labels=axes.get_xticks().tolist()
x_labels[0]='gogogo'
axes.set_xticklabels(x_labels,fontdict=font)
plt.show()

fig, axes = plt.subplots(1,2,squeeze=False)
axes[0,0].bar(x=[1,2,3],height=[4,5,6])
axes[0,1].bar(x=[4,5,6],height=[7,8,9])
ax0 = axes[0,0].get_xticks().tolist()
ax1 = axes[0,1].get_xticks().tolist()
ax0[2] = 'hello'
ax1[2] = 'world'
axes[0,0].set_xticklabels(ax0,fontdict=font)
axes[0,1].set_xticklabels(ax1,fontdict=font)
plt.show()