引言:
用python进行机器学习时需要分析数据画图及结果画图需要保存结果图片，为此，本博客稍微总结了一下常用的图片数据保存，如保存图像数据为pdf.
下面是一个用pytorch搭建的LSTM对sin函数进行预测，但是这不是本博客的重点，重点是总结一下图像数据保存，虽然内容小，但是对于像我这样的新手，显然不可或缺。
# -*- coding: utf-8 -*-
"""
Created on Sat May 4 18:38:06 2019
@author: adminster
"""
#机器学习或者深度学习，拟合程度可以非常高，但是要注意其泛化能力
from __future__ import print_function
import torch
import torch.nn as nn
import torch.optim as optim
import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt
class Sequence(nn.Module): #序列模型，名字能理解
def __init__(self):
super(Sequence, self).__init__()
self.lstm1 = nn.LSTMCell(1, 51)
self.lstm2 = nn.LSTMCell(51, 51)
self.linear = nn.Linear(51, 1)
def forward(self, input, future = 0):
outputs = []
h_t = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t = torch.zeros(input.size(0), 51, dtype=torch.double)
h_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
c_t2 = torch.zeros(input.size(0), 51, dtype=torch.double)
for i, input_t in enumerate(input.chunk(input.size(1), dim=1)):
print('i,input_t',i,input_t.shape)
h_t, c_t = self.lstm1(input_t, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))#承接关系
output = self.linear(h_t2)#h是记住的特征信息的状态
print('output',output.shape)
outputs += [output]
for i in range(future):# if we should predict the future
h_t, c_t = self.lstm1(output, (h_t, c_t))
h_t2, c_t2 = self.lstm2(h_t, (h_t2, c_t2))
output = self.linear(h_t2)
outputs += [output]
outputs = torch.stack(outputs, 1).squeeze(2)
return outputs
if __name__ == '__main__':
# set random seed to 0#查看之前结果的好方法
np.random.seed(0)
torch.manual_seed(0)
# load data and make training set #
data = torch.load('traindata.pt')
input = torch.from_numpy(data[3:, :-1])
#[-0.9613, -0.9738, -0.9840, ..., -0.7353, -0.7683, -0.7993]
target = torch.from_numpy(data[3:, 1:])
#[-0.9738, -0.9840, -0.9917, ..., -0.7683, -0.7993, -0.8283]
test_input = torch.from_numpy(data[:3, :-1])
test_target = torch.from_numpy(data[:3, 1:])
# build the model
seq = Sequence()
seq.double()#double()函数来自于object类对象
criterion = nn.MSELoss()#
# use LBFGS as optimizer since we can load the whole data to train
optimizer = optim.LBFGS(seq.parameters(), lr=0.8)
''' paramters()函数
for name, param in self.named_parameters():
yield param
'''
#begin to train
for i in range(15):
print('STEP: ', i)
def closure():
optimizer.zero_grad()
out = seq(input)#因为权重是随机的，当然就会是很错误的数据
loss = criterion(out, target)
print('loss:', loss.item())
loss.backward()
return loss
optimizer.step(closure)
# begin to predict, no need to track gradient here
with torch.no_grad():
future = 1000
pred = seq(test_input, future=future)#用训练好了的网络去预测
loss = criterion(pred[:, :-future], test_target)
print('test loss:', loss.item())
y = pred.detach().numpy()
# draw the result
plt.figure(figsize=(30,10))
plt.title('Predict future values for time sequences\n(Dashlines are predicted values)', fontsize=30)
plt.xlabel('x', fontsize=20)
plt.ylabel('y', fontsize=20)
plt.xticks(fontsize=20)
plt.yticks(fontsize=20)
def draw(yi, color):
plt.plot(np.arange(input.size(1)), yi[:input.size(1)], color, linewidth = 2.0)
plt.plot(np.arange(input.size(1), input.size(1) + future), yi[input.size(1):], color + ':', linewidth = 2.0)
draw(y[0], 'r')
draw(y[1], 'g')
draw(y[2], 'b')
#加一个文件夹
import time ,os #获取日期
time1=time.strftime('%Y-%m-%d')
sv_path='pre_data/'+time1
os.makedirs(sv_path,exist_ok=True)
plt.savefig(f'{sv_path}/predict%d.pdf'%i)#保存文件在指定文件夹下很方便
plt.close() #加一个文件夹
import time ,os #获取日期
time1=time.strftime('%Y-%m-%d')
sv_path='pre_data/'+time1
os.makedirs(sv_path,exist_ok=True)
plt.savefig(f'{sv_path}/predict%d.pdf'%i)#保存文件在指定文件夹下很方便
plt.close()
主要讲解一下这么一小段代码，高手轻拍，主要引入了两个包，time,os分别用来获取当前时间和创建文件夹，保存数据时这两步是必要操作，sv_path='pre_data/'+time1 用于获取年月日，
sv_path='pre_data/'+time1
os.makedirs(sv_path,exist_ok=True)创建多层文件夹目录，用于存放数据
plt.savefig(f'{sv_path}/predict%d.pdf'%i)#保存文件在指定文件夹下很方便
plt.close()调用matplotlib库中已有的函数savefig(path).里面可以保存tiff/png/jpg...等格式
