深度学习笔记

import torch
import os
import pandas as pd

#        第一部分
"""
x = torch.arange(12)
print(x)
print(x.shape)
print(x.numel())
X = x.reshape(3, 4)
print(X)
print(X.shape)
y = torch.zeros(2, 3, 4)
z = torch.ones(2, 3, 4)
print("y:", y, y.shape)
print("Z:", z, z.shape)
L = torch.tensor([[2, 1, 4, 3], [1, 2, 3, 4], [5, 6, 5, 4]])
print("L:", L, L.shape)

XX = torch.tensor([1.0, 2, 4, 8])
YY = torch.tensor([2, 2, 2, 2])
"""
"""加减乘除次方"""
# print(XX+YY, '\n', XX-YY, '\n', XX*YY, '\n', XX/YY, '\n',  XX**YY)

# x = torch.arange(12, dtype=torch.float32).reshape((3, 4))
# y = torch.tensor([[2.0, 1, 4, 3], [1, 2, 3, 4], [4, 3, 2, 1]])
"""cat将多个张量通过dim选择维度进行联结"""

"""
print(torch.cat((x, y), dim=0))
tensor([[ 0.,  1.,  2.,  3.],
        [ 4.,  5.,  6.,  7.],
        [ 8.,  9., 10., 11.],
        [ 2.,  1.,  4.,  3.],
        [ 1.,  2.,  3.,  4.],
        [ 4.,  3.,  2.,  1.]])
"""

"""
print(torch.cat((x, y), dim=1))
tensor([[ 0.,  1.,  2.,  3.,  2.,  1.,  4.,  3.],
        [ 4.,  5.,  6.,  7.,  1.,  2.,  3.,  4.],
        [ 8.,  9., 10., 11.,  4.,  3.,  2.,  1.]])
"""

"""
逻辑判断生成同型的bool矩阵
print(x == y)
tensor([[False,  True, False,  True],
        [False, False, False, False],
        [False, False, False, False]])
"""

"""
求和
print(x.sum())
tensor(66.)
"""
"""
x[-1]   取x的最后一行
x[1:3]  取x的2，3行,前闭后开,从0开始
x[1,2]  取x的第二行第三列
x[0:2, :]   0到1行，全部列  x:y，x可取，y取不到

id(x)   取x的地址，类似c中的&取址

格式转换
a = torch.tensor([3.5])

a           tensor([3.5000])
a.item()    3.5
float(a)    3.5
int(a)      3
"""

#       第二部分
# 文件夹在pycharm目录下
# os.mkdir(os.path.join('..', 'data'))
# data_file = os.path.join('..', 'data', 'house_tiny.csv')
# with open(data_file, 'w') as f:
#     f.write('NumRooms,Alley,Price\n')
#     f.write('NA,Pave,127500\n')
#     f.write('2,NA,10600\n')
#     f.write('4,NA,178100\n')
#     f.write('NA,NA,140000\n')
# data = pd.read_csv(data_file)
# print(data)
"""
data

   NumRooms Alley   Price
0       NaN  Pave  127500
1       2.0   NaN  106000
2       4.0   NaN  178100
3       NaN   NaN  140000
"""

# 处理缺失数据----插值法

# inputs, outputs = data.iloc[:, 0:2], data.iloc[:, 2]
# inputs = inputs.fillna(inputs.mean(numeric_only=True))
# print(inputs)
"""
inputs

   NumRooms Alley
0       3.0  Pave
1       2.0   NaN
2       4.0   NaN
3       3.0   NaN
"""

# inputs = pd.get_dummies(inputs, dummy_na=True, dtype=float)
# print(inputs)
"""
   NumRooms  Alley_Pave  Alley_nan
0       3.0        True      False
1       2.0       False       True
2       4.0       False       True
3       3.0       False       True

通过dtype=float
将True和False转为0.1
   NumRooms  Alley_Pave  Alley_nan
0       3.0         1.0        0.0
1       2.0         0.0        1.0
2       4.0         0.0        1.0
3       3.0         0.0        1.0

"""

# x, y = torch.tensor(inputs.values), torch.tensor(outputs.values)
# print(x)
# print(y)
"""
tensor([[3., 1., 0.],
        [2., 0., 1.],
        [4., 0., 1.],
        [3., 0., 1.]], dtype=torch.float64)
tensor([127500, 106000, 178100, 140000])
"""

# 要存梯度
x = torch.arange(4.0, requires_grad=True)
print("X:", x)
# y关于x的梯度是x.grad
print("求导前，y关于x的倒数:", x.grad)
y = 2 * torch.dot(x, x)     # 二倍的x的内积,即2*x^2,求导为4x
y.backward()
print("求导后，y关于x的倒数:", x.grad)

print(x.grad == x * 4)
x.grad.zero_()
y = x.sum()     # 即y=x1+x2+x3+...+xn,对每一项求偏导即为[1， 1， 1， 1]
y.backward()
print(x.grad)