- 折線圖
plt.figure(figsize=(40, 40)) # 確定圖像畫布的大小 plt.subplot(211) # 將畫布分為兩行一列 plt.xlabel('Number of sample', fontsize=40) # x軸的label plt.ylabel('Characteristics of the amplitude', fontsize=40) # y軸的label 備注(plot所有的原件都可以加fontsize屬性) plt.title('{} characteristics (ml_id=2 waveType=2)'.format(c_type), fontsize=50) # 圖的title plt.plot(two_type_list[:two_negative_end_index], linestyle = "-", color = 'r', # 繪制折線圖,其中若x參數省略,則橫坐標以y列表的索引代替 label = 'Negative | average: {} variance: {} median: {}'.format(('%.2f' % np.mean(two_type_list[ : two_negative_end_index])), # label參數表示這條線的label,可以當作圖例顯示出來 ('%.2f' % np.var(two_type_list[ : two_negative_end_index])), ('%.2f' % np.median(two_type_list[ : two_negative_end_index]))), linewidth=3.0) # 線寬 plt.plot(two_type_list[two_negative_end_index+1:], linestyle = "-", color = 'g', # 備注(一張圖可以累積加多個plot) label = 'Positive | average: {} variance: {} median: {}'.format(('%.2f' % np.mean(two_type_list[two_negative_end_index+1 : ])), ('%.2f' % np.var(two_type_list[two_negative_end_index+1 : ])), ('%.2f' % np.median(two_type_list[two_negative_end_index+1 : ]))), linewidth=3.0) # plt.ylim(0, 5) # 設置y軸的取值范圍,如設置(0,5)則y軸坐標為從0開始,到5結束 # 刻度值字體大小設置 plt.tick_params(labelsize=40) # 設置坐標軸上刻度的字體大小 plt.legend(loc=0, fontsize = 40) # 顯示圖例,loc=0表示圖例會根據圖片情況自動擺放 #################################################################################################################################### plt.subplot(212) plt.xlabel('Number of sample', fontsize=40) plt.ylabel('Characteristics of the amplitude', fontsize=40) plt.title('{} characteristics (ml_id=6 waveType=2)'.format(c_type), fontsize=50) plt.plot(six_type_list[:six_negative_end_index], linestyle = "-", color = 'r', label = 'Negative | average: {} variance: {} median: {}'.format(('%.2f' % np.mean(six_type_list[ : six_negative_end_index])), ('%.2f' % np.var(six_type_list[ : six_negative_end_index])), ('%.2f' % np.median(six_type_list[ : six_negative_end_index]))), linewidth=3.0) plt.plot(six_type_list[six_negative_end_index+1:], linestyle = "-", color = 'g', label = 'Positive | average: {} variance: {} median: {}'.format(('%.2f' % np.mean(six_type_list[six_negative_end_index+1 : ])), ('%.2f' % np.var(six_type_list[six_negative_end_index+1 : ])), ('%.2f' % np.median(six_type_list[six_negative_end_index+1 : ]))), linewidth=3.0) # 刻度值字體大小設置 plt.tick_params(labelsize=40) plt.legend(loc=0, fontsize = 40) plt.savefig('C:/Users/Mloong/Desktop/f_image/{} characteristics.png'.format(c_type), dpi=300) plt.show()
2.散點圖
_type = 'median' plt.scatter(range(0, 3790), two_avgAbs_list[0:3790], c='r') # 散點圖的x參數不可省略 plt.scatter(range(3791, 4939), two_avgAbs_list[3791:4939], c='g') plt.title('{} ml_id=2 waveType=2'.format(_type)) plt.savefig('C:/Users/Mloong/Desktop/f_image/{} scatter ml_id=2 waveType=2.png'.format(_type), dpi=300) plt.show()
3.概率分布圖
# 概率分布圖 type_list = two_median_list _type = 'median' num_bins = 100 # 條狀圖的個數 plt.hist(type_list[:3790], num_bins, normed=1, facecolor='blue', alpha=0.5) plt.hist(type_list[3791:], num_bins, normed=1, facecolor='red', alpha=0.5) plt.xlabel('Value') plt.ylabel('Probability') plt.title('{} probability distribution ml_id=2 waveType=2'.format(_type)) plt.subplots_adjust(left=0.15) plt.savefig('C:/Users/Mloong/Desktop/f_image/{} probability distribution ml_id=2 waveType=2.png'.format(_type), dpi=300) plt.show()
4.箱形圖
_type = 'pca_value' import seaborn as sns plt.subplot(121) plt.title('{} (ml_id=2 waveType=2)'.format(_type)) sns.set(style='whitegrid') # 設置背景 sns.boxplot(x='label', y='{}'.format(_type), data=two_data, hue='label') # data參數是一個dataframe對象,其中x和y分別時這個dataframe中的列名 ######################################################################################### plt.subplot(122) plt.title('{} (ml_id=6 waveType=2)'.format(_type)) sns.set(style='whitegrid') # 設置背景 sns.boxplot(x='label', y='{}'.format(_type), data=six_data, hue='label') # 繪制箱形圖 plt.savefig('C:/Users/Mloong/Desktop/f_image/{} box figure.png'.format(_type), dpi=300) plt.show()
5.熱圖
# 2.相關矩陣 import seaborn as sns corrmat = two_data[['avs', 'avgAbs', 'rms', 'rms2', 'wave', 'pulse', 'PeekFlag', 'Margin', 'Skewness', 'Kurtosis', 'median', 'pca_value', 'label']].corr() # .corr()求相關矩陣,此時返回的值corrmat為相關矩陣 f, ax = plt.subplots(figsize=(12, 9)) sns.heatmap(corrmat, vmax=.8, square=True) # 將這個相關矩陣以熱圖的形式畫出來 plt.savefig('C:/Users/Mloong/Desktop/f_image/two correlation matrix.png', dpi=300) plt.show()