快捷導(dǎo)航

Python使用matplotlib.pyplot畫(huà)熱圖和損失圖的代碼詳解

更新時(shí)間：2023年09月19日 10:53:23 作者：童話(huà)ing

眾所周知,在完成論文相關(guān)工作時(shí)畫(huà)圖必不可少,如損失函數(shù)圖、熱力圖等是非常常見(jiàn)的圖,在本文中,總結(jié)了這兩個(gè)圖的畫(huà)法,下面給出了完整的代碼,開(kāi)箱即用,感興趣的同學(xué)可以自己動(dòng)手嘗試一下

一、損失函數(shù)圖

import matplotlib.pyplot as plt
file = open('E:\\5120154230PythonCode\\PBAN-PyTorch-master\\state_dict\\loss\\PBAN_New_restaurant15_0.001_80_0.2_16.csv')  # 打開(kāi)文檔
data = file.readlines()  # 讀取文檔數(shù)據(jù)
para_1 = []  # 新建列表，用于保存第一列數(shù)據(jù)
para_2 = []  # 新建列表，用于保存第二列數(shù)據(jù)
cnt = 0
for num in data:
    try:
        temp = num.split(",")
        cnt += 1
        if cnt==700:
            break
    except:
        continue
    para_1.append(float(num.split(',')[0]))
    para_2.append(float(num.split(',')[1]))
plt.figure()
# plt.title('loss')
plt.xlabel("iterations")
plt.ylabel("loss")
#color in cnblogs.com/qccc/p/12795541.html
#orange、teal、red、chocolate
plt.plot(para_1, para_2)
plt.show()

CSV數(shù)據(jù)格式：第一列為Epoch或者迭代次數(shù)等，第二列為損失值。

效果圖：

二、熱圖

import matplotlib.pyplot as plt
import pandas as pd
import matplotlib.ticker as ticker
d = [
 [0.43757705, 0.30564879, 0.08757705, 0.013755781, 0.13755781, 0.04080211, 0.03615228],
 [0.31525328, 0.42328909, 0.04004493, 0.01735733, 0.01755249, 0.02630009, 0.09020273],
 [0.01546572, 0.09022246, 0.4166335, 0.09773314, 0.10259592, 0.0447391, 0.03261019],
 [0.01536734, 0.010553601, 0.045800883, 0.39755909, 0.1465714, 0.0408309, 0.03612638],
 [0.11513351, 0.01193435, 0.051866556, 0.046714543, 0.42510962, 0.03154159, 0.4848393],
 [0.11544053, 0.0941444, 0.050161916, 0.09768857, 0.11385846, 0.43073818, 0.13351071],
 [0.01529034, 0.07752335, 0.04121181, 0.01742287, 0.35099512, 0.03777161, 0.38087882]
]
variables = ['Great', 'food', 'but', 'the', 'service', 'was', 'dreadful']
labels = ['Great', 'food', 'but', 'the', 'service', 'was', 'dreadful']
df = pd.DataFrame(d, columns=variables, index=labels)
fig = plt.figure(figsize=(7, 6)) #寬、高
ax = fig.add_subplot(1, 1, 1) #畫(huà)布設(shè)置為1行1列顯示在第一塊中
# cmap參考：https://matplotlib.org/2.0.2/users/colormaps.html
# hot_r、afmhot_r、plasma_r、ocean_r
# interpolation：nearest，None、none
cax = ax.matshow(df, interpolation='nearest', cmap='hot_r')
fig.colorbar(cax)
tick_spacing = 1
ax.xaxis.set_major_locator(ticker.MultipleLocator(tick_spacing))
ax.yaxis.set_major_locator(ticker.MultipleLocator(tick_spacing))
ax.set_xticklabels([''] + list(df.columns))
ax.set_yticklabels([''] + list(df.index))
plt.show()

效果：

另一份代碼：

import matplotlib.pylab as plt
import numpy as np
def samplemat(dims):
    aa = np.zeros(dims)
    for i in range(dims[1]):
            aa[0,i] = i
    return aa
dimlist = [(1, 12)]
for d in dimlist:
    arr = samplemat(d)
    plt.matshow(arr)
plt.show()