Stephanie的博客

1.算频谱def spectrum_plot(signal,fs,titlename): ''' Parameters ---------- signal : numpy fs : GBaud titlename : str Returns ------- ''' if isinstance(signal,torch.Tensor): signal=signal.cpu().detach().nump

有两种方法完成数据预处理，一种是写循环，另一种是生成索引矩阵，然后索引，下面来看下时间# -*- coding: utf-8 -*-"""Created on Tue Jul 13 10:55:19 2021在神经网络中将序列转换为滑窗矩阵的算法时间对比@author: user"""import timeimport numpy as npimport torchimport matplotlib.pyplot as pltdef window_stack(a,stepsi

import randomxx=[random.uniform(1,10) for i in range(5)]max_xx=max(xx)max_index_xx=xx.index(max(xx))

我要看懂这个代码！！import numpy as npimport tensorflow as tfimport matplotlib.pyplot as pltnp.random.seed(0) # for reproducibilityX = np.random.randint(0, 3, (32*1100, 10), np.int32) # inputsY = np.equal(np.concatenate((np.ones_like(X[:, :1]), np.sign(X[:

1.np.selectnp.select(condlist, choicelist, default=0)condlist参数：操作数据所依据的条件choicelist参数：根据condlist条件，索要执行的操作返回值：返回的是一个“ 列表” 。注意上面的condlist和choicelist都必须是写成“ 列表 ”的形式。...

import random#生成1到10之间的均匀分布的一个数random.uniform(1,10)#正太分布 normalvariate(mu,sigma) mu 均值 sigma 标准差#mu=0 sigma=1 标准正太分布 标准差是指此范围所占比率为全部数值之68%random.normalvariate(0,1) ...

1.numpy升维比如维度由(5,)变为(5,1)import numpy as npa=np.random.randn(5)b=np.expand_dims(a,axis=1)print(a.shape)print(b.shape)输出2.降维c=b.squeeze()print(c.shape)输出

保存，保存list,numpy,torch的类型都行import numpy as npimport torchimport scipy.io as siovectest=[1,2,3,4]vectest2=np.array(vectest)vectest3=torch.tensor(vectest)path='test.mat'sio.savemat(path,{"v":vectest})

import numpy as npwith open('eye.txt','ab') as f: newresult1 = np.random.rand(2, 3) newresult2 = np.random.rand(2, 3) np.savetxt(f, newresult1, delimiter=" ") np.savetxt(f, newresult2, delimiter=" ") with open('eye.txt','ab') as f: .