目录

1.超过二维的张量写进csv

2.翻转矩阵与绘制热力图

3.切片

4.np按块复制

---

1.超过二维的张量写进csv

#(20,204,273) -> (4080,273)
ycsv = []

for i in range(20):
    ycsv.append(y[i, 8, :, :].reshape(204,273))

with open('y.csv','w',encoding='utf-8') as y_obj:
    writer = csv.writer(y_obj)
    for j in range(20):
        writer.writerows(ycsv[j])

data = pd.read_csv('y.csv',encoding='utf-8')
print(data.shape)

2.翻转矩阵与绘制热力图

#上下翻转矩阵
y = y[::-1,:]

#度数符号
du = chr(176)

#vmax设置颜色条上界，vmin设置下界，cmap设置颜色条样例
ax1 = sns.heatmap(predy,cmap = 'hot_r',vmax = 160,vmin = 0)

#xticks中可以用ticks来设置labels打点位置，fontsize设置字体大小，rotation设置旋转角度
plt.xticks(ticks=[0,34,68,102,136,170,204,238,272],labels=[str(x)+du+'E' for x in range(109,118,1)],fontsize=10,rotation=360)

plt.yticks(ticks=[29,64,99,134,169,204], labels=[str(x)+du+'N' for x in [25,24,23,22,21,20]],fontsize=10)

#设置labels与图片的距离
plt.tick_params(axis='x',pad=5)

plt.tick_params(axis='y',pad=5)

figure1 = ax1.get_figure()

figure1.savefig('predy.jpg')

3.切片

import numpy as np

a = [0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]
b = a[:5]
c = a[6:9]  #前面六个省略，从第七个开始取三个，左开右闭
d = a[9:15]  #连续读取就顶着前面的继续切
b, c, d

4.np按块复制

import numpy as np

a = np.zeros((24,40,40,70))
b = np.tile(a,(30,1,1,1))
c = np.tile(a,(1,30,1,1))