一、概念：张量，算子
张量（Tensor）：和numpy类似，张量就是一个多维数组。
张量的维度和数组的维度一样。
如下图所示：
举例：RGB就是一个三维图像，RGB中的每一个像素为一个维度。
Tensor有三个属性：
rank:number of dimensions(维度的数目)
shape:number of rows and columns(行和列的数目)
type: data type of tensor’s elements（数据的类型）
numpy和tensor的联系：二者共享内存，numpy和tensor之间可以进行相互转化。
区别：
1、numpy 默认类型是 float64、int32，tensor 默认类型是float32、int64
2、numpy只支持cpu加速，tensor不仅支持cou加速还支持gpu加速，很牛。
算子
人工智能的算子实际是一定的函数。可以对函数的运算进行定义。说白了就是一种算法.T_T.
二、张量
1.2张量
1.2.1创建张量
生成张量的numpy数组==>转换为对应的张量
1.2.1.1 指定数据创建张量
输出数组生成Numpy对象==>转换为tensor张量。
import torch
import numpy as np
#创建一个数组num1
num_1 = np.array([2.3,3.3,4.0]);
#将num1转换为tensor数据
num_1_to_tensor = torch.tensor(num_1);
print(num_1_to_tensor)
'''
2,3,4为输入的数据
tensor([2.3000, 3.3000, 4.0000], dtype=torch.float64)
'''
1.2.1.2 指定形状创建
生成指定形状的numpy数组==>转换为tensor张量
#创建大小为三行四列的一个张量
num_2 = torch.ones([3,4])
print(num_2)
'''
tensor([[1., 1., 1., 1.],
[1., 1., 1., 1.],
[1., 1., 1., 1.]])
'''
1.2.1.3 指定区间创建
生成指定区间的numpy数组==>转换为tensor张量
import torch
import numpy as np
#指定区间创建张量
arrange_tensor = torch.arange(start=1,end=5,step=1); #1开始，5结束，但是不包括5.
linespace_tensor = torch.linspace(start=1,end=5,steps=5);#1开始，5结束，包括5，steps为平分成几个数，
print('arange Tensor:',arrange_tensor);
print('linespace Tensor:',linespace_tensor);
'''
arange Tensor: tensor([1, 2, 3, 4])
linespace_steps为5的情况下
linespace Tensor: tensor([1., 2., 3., 4., 5.])
line_space_steps为6的情况下
linespace Tensor: tensor([1.0000, 1.8000, 2.6000, 3.4000, 4.2000, 5.0000])
'''
1.2.2张量的属性
1.2.2.1张量的形状
import torch
import numpy as np
#指定区间创建张量
#生成一个2*3*4*5维度大小的张量
ndim_4_Tensor = torch.ones([2,3,4,5])
#输出向量的维度
print("Number of dimensions:", ndim_4_Tensor.ndim)
#输出向量的大小
print("Shape of Tensor:", ndim_4_Tensor.shape)
#输出第0维的大小
print("Elements number along axis 0 of Tensor:", ndim_4_Tensor.shape[0])
#输出最后一维的大小
print("Elements number along the last axis of Tensor:", ndim_4_Tensor.shape[-1])
#输出总数
print('Number of elements in Tensor: ', ndim_4_Tensor.size)
'''
Number of dimensions: 4
Shape of Tensor: torch.Size([2, 3, 4, 5])
Elements number along axis 0 of Tensor: 2
Elements number along the last axis of Tensor: 5
Number of elements in Tensor: <built-in method size of Tensor object at 0x000001C79A6996D8>
'''
1.2.2.2 形状的改变
#生成数组
num3_array = np.array([[[1, 2, 3, 4, 5],
[6, 7, 8, 9, 10]],
[[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20]],
[[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30]]]);
num3_tensor = torch.tensor(num3_array);
#改变tensor张量的形状,但是总数最后要求一致
#举例：3*2*5 === 2*3*5
num3_tensor_reshape = torch.reshape(num3_tensor,[2,3,5])
print('Before reshape:',num3_tensor.shape);
print('After reshape:',num3_tensor_reshape.shape);
'''
Before reshape: torch.Size([3, 2, 5])
tensor([[[ 1, 2, 3, 4, 5],
[ 6, 7, 8, 9, 10]],
[[11, 12, 13, 14, 15],
[16, 17, 18, 19, 20]],
[[21, 22, 23, 24, 25],
[26, 27, 28, 29, 30]]], dtype=torch.int32)
After reshape: torch.Size([2, 3, 5])
tensor([[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
[16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]],
dtype=torch.int32)
'''
#也可以变成2维的
#2*15 === 3*2*5 === 2*3*5
num3_tensor_reshape_to_2dim = torch.reshape(num3_tensor,[2,15])
print('After reshape to 2 dim:',num3_tensor_reshape_to_2dim.shape)
'''
After reshape to 2 dim: torch.Size([2, 15])
tensor([[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
[16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]],
dtype=torch.int32)
'''
改变张量的形状，我自己的理解昂，就是先把张量变成一维数组，在从第一维开始排，一直排到最后一维。
1.2.2.3 张量的数据类型
张量生成中，如果咱们没有指定数据类型的话。
整数会生成int 64。浮点数会生成float 32
print("int type",torch.tensor(1).dtype)
print("float type",torch.tensor(1.0).dtype)
'''
int type torch.int64
float type torch.float32
'''
#在pytorch官方文档中并没有找到转换类型的函数，类似于飞桨的cast，但可以在生成的时候就改变他的类型
print("int to float",torch.ones([2,2],dtype=torch.float32))
'''
int to float
tensor([[1., 1.],
[1., 1.]])
'''
1.2.2.4 张量的设备位置
垃圾电脑没安装gpu版本的pytorch，这里建议大家登录kaggle官网，上面由免费时间的gpu加速，嘎嘎快！！
import numpy as np
import torch
#生成指定数组
num1 = np.array([[1,2,3],[4,5,6]]);
#转换为张量并存储在cpu上
num1_cpu = torch.tensor(num1,device=torch.device("cpu"))
print("The position of num1_tensor(CPU):",num1_cpu.device)
#转换为张量并存储在GPU上
num1_gpu = torch.tensor(num1,device=torch.device("cuda"))
print("The position of num1_tensor(GPU):",num1_gpu.device)
'''
The position of num1_tensor(CPU): cpu
The position of num1_tensor(GPU): cuda:0
AIStudio上介绍的时候有一种固定内存，我在搜索torch.devcie参数的时候并没有找到指定固定内存的参数，所以只生成了在gpu和cpu设备上的张量，由于我的电脑真的垃圾，真烦人,没有gpu版本的torch,是在kaggle官网上利用gpu运行的。
'''
1.2.3 张量与Numpy数组转换（张量和numpy数组之间是可以相互转换）
#numpy数组转换位张量
num_numpy = np.array([[1,2,3],[4,5,6]])
num_numpy_to_tensor = torch.tensor(num_numpy);
print('numpy convert to tensor(定义)',num_numpy_to_tensor)
#也可以使用函数torch.from_numpy
print('numpy convert to tensor(函数)',torch.from_numpy(num_numpy))
#将张量转换为tensor数组，就利用上面的在转换回去
print('tensor convert to numpy',num_numpy_to_tensor.numpy())
'''
numpy convert to tensor(定义) :
tensor([[1, 2, 3],
[4, 5, 6]], dtype=torch.int32)
numpy convert to tensor(函数) :
tensor([[1, 2, 3],
[4, 5, 6]], dtype=torch.int32)
tensor convert to numpy :
[[1 2 3]
[4 5 6]]
'''
1.2.4张量的访问
1.2.4.1 索引和切片
#生成张量
num_numpy = np.array([[1,2,3],[4,5,6]])
num_numpy_to_tensor = torch.tensor(num_numpy);
#索引:和普通的数组获取索引一样索引,对应的下标
for i in range(0, len(num_numpy_to_tensor)):
print(num_numpy_to_tensor[i])
#切片：和普通的数组切片一样,start stop step
print(num_numpy_to_tensor[0:1:])
'''
tensor([1, 2, 3], dtype=torch.int32)
tensor([4, 5, 6], dtype=torch.int32)
tensor([[1, 2, 3]], dtype=torch.int32)
'''
1.2.4.2 访问张量
#对张量进行访问
num_numpy = np.array([[1,2,3],[4,5,6]])
num_numpy_to_tensor = torch.tensor(num_numpy);
print("访问第一维:",num_numpy_to_tensor[0])
print("访问第一维第二个数字",num_numpy_to_tensor[1,1])
print("访问第一维的所有数字",num_numpy_to_tensor[1:])
'''
访问第一维: tensor([1, 2, 3], dtype=torch.int32)
访问第一维第二个数字:tensor(5, dtype=torch.int32)
访问第一维的所有数字:tensor([[4, 5, 6]], dtype=torch.int32)
'''
1.2.4.3 修改张量
#改变张量中某一个具体的数
print('before change',num_numpy_to_tensor)
#改变第二个的第二个数为2
num_numpy_to_tensor[1,1]=2;
print('after change',num_numpy_to_tensor)
'''
before change tensor([[1, 2, 3],
[4, 5, 6]], dtype=torch.int32)
after change tensor([[1, 2, 3],
[4, 2, 6]], dtype=torch.int32)
'''
1.2.5 张量的运算
1.2.5.1 数学运算
num1 = np.array([[1,-1,2],[2,-3,4.1]]);
num1_tensor = torch.tensor(num1);
#转换为绝对值
print('取绝对值abs:',num1_tensor.abs());
print('挨个向上取整:',num1_tensor.floor());
print('挨个四舍五入:',num1_tensor.round())
print('挨个元素计算自然常数为底的指数:',num1_tensor.exp())
print('挨个取自然对数:',num1_tensor.log())
print('逐个元素取倒数:',num1_tensor.reciprocal())
print('逐个元素计算平方',num1_tensor.square())
print('逐个元素计算平方根',num1_tensor.sqrt())
print('逐个元素计算正弦',num1_tensor.sin())
print('逐个元素计算余弦',num1_tensor.cos())
print('逐个元素相加',num1_tensor.add(1))
print('逐个元素相减',num1_tensor.subtract(1))
print('逐个元素相乘',num1_tensor.multiply(2))
print('逐个元素相除',num1_tensor.divide(2))
print('逐个元素取幂',num1_tensor.pow(2))
'''
取绝对值abs: tensor([[1.0000, 1.0000, 2.0000],
[2.0000, 3.0000, 4.1000]], dtype=torch.float64)
挨个向上取整: tensor([[ 1., -1., 2.],
[ 2., -3., 4.]], dtype=torch.float64)
挨个四舍五入: tensor([[ 1., -1., 2.],
[ 2., -3., 4.]], dtype=torch.float64)
挨个元素计算自然常数为底的指数: tensor([[2.7183e+00, 3.6788e-01, 7.3891e+00],
[7.3891e+00, 4.9787e-02, 6.0340e+01]], dtype=torch.float64)
挨个取自然对数: tensor([[0.0000, nan, 0.6931],
[0.6931, nan, 1.4110]], dtype=torch.float64)
逐个元素取倒数: tensor([[ 1.0000, -1.0000, 0.5000],
[ 0.5000, -0.3333, 0.2439]], dtype=torch.float64)
逐个元素计算平方 tensor([[ 1.0000, 1.0000, 4.0000],
[ 4.0000, 9.0000, 16.8100]], dtype=torch.float64)
逐个元素计算平方根 tensor([[1.0000, nan, 1.4142],
[1.4142, nan, 2.0248]], dtype=torch.float64)
逐个元素计算正弦 tensor([[ 0.8415, -0.8415, 0.9093],
[ 0.9093, -0.1411, -0.8183]], dtype=torch.float64)
逐个元素计算余弦 tensor([[ 0.5403, 0.5403, -0.4161],
[-0.4161, -0.9900, -0.5748]], dtype=torch.float64)
逐个元素相加 tensor([[ 2.0000, 0.0000, 3.0000],
[ 3.0000, -2.0000, 5.1000]], dtype=torch.float64)
逐个元素相减 tensor([[ 0.0000, -2.0000, 1.0000],
[ 1.0000, -4.0000, 3.1000]], dtype=torch.float64)
逐个元素相乘 tensor([[ 2.0000, -2.0000, 4.0000],
[ 4.0000, -6.0000, 8.2000]], dtype=torch.float64)
逐个元素相除 tensor([[ 0.5000, -0.5000, 1.0000],
[ 1.0000, -1.5000, 2.0500]], dtype=torch.float64)
逐个元素取幂 tensor([[ 1.0000, 1.0000, 4.0000],
[ 4.0000, 9.0000, 16.8100]], dtype=torch.float64)
Process finished with exit code 0
'''
1.2.5.2 逻辑运算
num_numpy = np.array([[1,2,3],[4,5,6]])
num_numpy_to_tensor = torch.tensor(num_numpy);
print('是否为有限',num_numpy_to_tensor.isfinite())
# 判断两个Tensor的每个元素是否相等，并返回形状相同的布尔类Tensor
print("判断两个张量的每个元素是否相等",num_numpy_to_tensor.equal(num_numpy_to_tensor))
print("判断两个张量的每个元是否不相等",num_numpy_to_tensor.not_equal(num_numpy_to_tensor))
print("判断x的元素是否小于y的元素",num_numpy_to_tensor.less(num_numpy_to_tensor))
print("判断x的元素是否小于或等于y的元素",num_numpy_to_tensor.less_equal(num_numpy_to_tensor))
print("判断x的元素是否大于y的对应元素",num_numpy_to_tensor.greater(num_numpy_to_tensor))
print("判断x的元素是否大于等于y的元素",num_numpy_to_tensor.greater_equal(num_numpy_to_tensor))
print("判断元素是否接近",num_numpy_to_tensor.allclose(num_numpy_to_tensor))
"""x.less_than(y) # 判断Tensor x的元素是否小于Tensor y的对应元素
'''
是否为有限 tensor([[True, True, True],
[True, True, True]])
判断两个张量的每个元素是否相等 True
判断两个张量的每个元是否不相等 tensor([[False, False, False],
[False, False, False]])
判断x的元素是否小于y的元素 tensor([[False, False, False],
[False, False, False]])
判断x的元素是否小于或等于y的元素 tensor([[True, True, True],
[True, True, True]])
判断x的元素是否大于y的对应元素 tensor([[False, False, False],
[False, False, False]])
判断x的元素是否大于等于y的元素 tensor([[True, True, True],
[True, True, True]])
判断元素是否接近 True
Process finished with exit code 0
'''
1.2.5.3 矩阵运算
num_numpy = np.array([[1,2.0,3],[4,5,6]])
num_numpy_to_tensor = torch.tensor(num_numpy);
#矩阵的转置
print(num_numpy_to_tensor.t())
#交换第0维和第1维
print(torch.transpose(num_numpy_to_tensor,1,0))
#矩阵的弗洛贝尼乌斯范数
print(torch.norm(num_numpy_to_tensor,p="fro"))
#矩阵(x-y)的2范数
print(num_numpy_to_tensor.dist(num_numpy_to_tensor,p=2))
#矩阵乘法
print(num_numpy_to_tensor.matmul(num_numpy_to_tensor.t()))
'''
tensor([[1., 4.],
[2., 5.],
[3., 6.]], dtype=torch.float64)
tensor([[1., 4.],
[2., 5.],
[3., 6.]], dtype=torch.float64)
tensor(9.5394, dtype=torch.float64)
tensor(0., dtype=torch.float64)
tensor([[14., 32.],
[32., 77.]], dtype=torch.float64)
'''
1.2.5.4 广播机制
x_numpy = np.ones((2,3,1,5))
y_numpy = np.ones((3,4,1))
x_tensor = torch.tensor(x_numpy)
y_tensor = torch.tensor(y_numpy)
z = x_tensor+y_tensor
print('broadcasting with two different shape tensor:',z.shape);
'''
broadcasting with two different shape tensor: torch.Size([2, 3, 4, 5])
'''
从输出结果看，此时x和y是不能广播的，因为在第一次从后往前的比较中，4和6不相等，不符合广播规则。
广播机制的计算规则
现在我们知道在什么情况下两个张量是可以广播的。两个张量进行广播后的结果张量的形状计算规则如下：
1）如果两个张量shape的长度不一致，那么需要在较小长度的shape前添加1，直到两个张量的形状长度相等。
2） 保证两个张量形状相等之后，每个维度上的结果维度就是当前维度上较大的那个。
以张量x和y进行广播为例，x的shape为[2, 3, 1，5]，张量y的shape为[3，4，1]。首先张量y的形状长度较小，因此要将该张量形状补齐为[1, 3, 4, 1]，再对两个张量的每一维进行比较。从第一维看，x在一维上的大小为2，y为1，因此，结果张量在第一维的大小为2。以此类推，对每一维进行比较，得到结果张量的形状为[2, 3, 4, 5]。
由于矩阵乘法函数paddle.matmul在深度学习中使用非常多，这里需要特别说明一下它的广播规则：
1）如果两个张量均为一维，则获得点积结果。
2） 如果两个张量都是二维的，则获得矩阵与矩阵的乘积。
3） 如果张量x是一维，y是二维，则将x的shape转换为[1, D]，与y进行矩阵相乘后再删除前置尺寸。
4） 如果张量x是二维，y是一维，则获得矩阵与向量的乘积。
5） 如果两个张量都是N维张量（N > 2），则根据广播规则广播非矩阵维度（除最后两个维度外其余维度）。比如：如果输入x是形状为[j,1,n,m]的张量，另一个y是[k,m,p]的张量，则输出张量的形状为[j,k,n,p]。
三. 使用pytorch实现数据预处理
1. 读取数据集 house_tiny.csv、boston_house_prices.csv、Iris.csv
import pandas as pd
import torch
#读取数据集
data_iris = pd.read_csv('Iris.csv')
data_house_tiny = pd.read_csv('house_tiny.csv')
data_boston_house_prices = pd.read_csv('boston_house_prices.csv')
#输出数据集
print(data_iris)
print(data_house_tiny)
print(data_boston_house_prices)
'''
Id SepalLengthCm ... PetalWidthCm Species
0 1 5.1 ... 0.2 Iris-setosa
1 2 4.9 ... 0.2 Iris-setosa
2 3 4.7 ... 0.2 Iris-setosa
3 4 4.6 ... 0.2 Iris-setosa
4 5 5.0 ... 0.2 Iris-setosa
.. ... ... ... ... ...
145 146 6.7 ... 2.3 Iris-virginica
146 147 6.3 ... 1.9 Iris-virginica
147 148 6.5 ... 2.0 Iris-virginica
148 149 6.2 ... 2.3 Iris-virginica
149 150 5.9 ... 1.8 Iris-virginica
[150 rows x 6 columns]
NumRooms Alley Price
0 NaN Pave 127500
1 2.0 NaN 106000
2 4.0 NaN 178100
3 NaN NaN 140000
CRIM ZN INDUS CHAS NOX ... RAD TAX PTRATIO LSTAT MEDV
0 0.00632 18.0 2.31 0 0.538 ... 1 296 15.3 4.98 24.0
1 0.02731 0.0 7.07 0 0.469 ... 2 242 17.8 9.14 21.6
2 0.02729 0.0 7.07 0 0.469 ... 2 242 17.8 4.03 34.7
3 0.03237 0.0 2.18 0 0.458 ... 3 222 18.7 2.94 33.4
4 0.06905 0.0 2.18 0 0.458 ... 3 222 18.7 5.33 36.2
.. ... ... ... ... ... ... ... ... ... ... ...
501 0.06263 0.0 11.93 0 0.573 ... 1 273 21.0 9.67 22.4
502 0.04527 0.0 11.93 0 0.573 ... 1 273 21.0 9.08 20.6
503 0.06076 0.0 11.93 0 0.573 ... 1 273 21.0 5.64 23.9
504 0.10959 0.0 11.93 0 0.573 ... 1 273 21.0 6.48 22.0
505 0.04741 0.0 11.93 0 0.573 ... 1 273 21.0 7.88 11.9
[506 rows x 13 columns]
Process finished with exit code 0
'''
- 处理缺失值
import pandas as pd
import torch
#读取数据集
data_iris = pd.read_csv('Iris.csv')
data_house_tiny = pd.read_csv('house_tiny.csv')
data_boston_house_prices = pd.read_csv('boston_house_prices.csv')
#输出数据集
print(data_iris)
print(data_house_tiny)
print(data_boston_house_prices)
#处理缺失值和离散值
input,output = data_iris.iloc[:,0:5],data_iris.iloc[:,5]
#使用均值处理缺失值
input = input.fillna(input.mean())
print(input)
input = np.array(input)
#处理离散值
output = pd.get_dummies(output,dummy_na=True)
output = np.array(output)
print(input,output)
#读取输入输出
input_house_tiny,output_house_tiny = data_house_tiny.iloc[:,0:2],data_house_tiny.iloc[:,2]
#处理缺失值
input_house_tiny = input_house_tiny.fillna(input_house_tiny.mean())
#处理离散值
input_house_tiny = pd.get_dummies(input_house_tiny,dummy_na=True)
input_house_tiny,output_house_tiny = np.array(input_house_tiny),np.array(output_house_tiny)
priint(input_house_tiny,output_house_tiny)
#读取输出
input_boston_house_prices,output_boston_house_prices = data_boston_house_prices.iloc[:,0:12],data_boston_house_prices.iloc[:,12]
#处理缺失值
input_boston_house_prices,output_boston_house_prices = input_boston_house_prices.fillna(input_boston_house_prices.mean()),output_boston_house_prices.fillna(output_boston_house_prices.mean())
print(input_boston_house_prices,output_boston_house_prices)
'''
[[1.00e+00 5.10e+00 3.50e+00 1.40e+00 2.00e-01]
[2.00e+00 4.90e+00 3.00e+00 1.40e+00 2.00e-01]
[3.00e+00 4.70e+00 3.20e+00 1.30e+00 2.00e-01]
[4.00e+00 4.60e+00 3.10e+00 1.50e+00 2.00e-01]
[5.00e+00 5.00e+00 3.60e+00 1.40e+00 2.00e-01]
[6.00e+00 5.40e+00 3.90e+00 1.70e+00 4.00e-01]
[7.00e+00 4.60e+00 3.40e+00 1.40e+00 3.00e-01]
[8.00e+00 5.00e+00 3.40e+00 1.50e+00 2.00e-01]
[9.00e+00 4.40e+00 2.90e+00 1.40e+00 2.00e-01]
[1.00e+01 4.90e+00 3.10e+00 1.50e+00 1.00e-01]
[1.10e+01 5.40e+00 3.70e+00 1.50e+00 2.00e-01]
[1.20e+01 4.80e+00 3.40e+00 1.60e+00 2.00e-01]
[1.30e+01 4.80e+00 3.00e+00 1.40e+00 1.00e-01]
[1.40e+01 4.30e+00 3.00e+00 1.10e+00 1.00e-01]
[1.50e+01 5.80e+00 4.00e+00 1.20e+00 2.00e-01]
[1.60e+01 5.70e+00 4.40e+00 1.50e+00 4.00e-01]
[1.70e+01 5.40e+00 3.90e+00 1.30e+00 4.00e-01]
[1.80e+01 5.10e+00 3.50e+00 1.40e+00 3.00e-01]
[1.90e+01 5.70e+00 3.80e+00 1.70e+00 3.00e-01]
[2.00e+01 5.10e+00 3.80e+00 1.50e+00 3.00e-01]
[2.10e+01 5.40e+00 3.40e+00 1.70e+00 2.00e-01]
[2.20e+01 5.10e+00 3.70e+00 1.50e+00 4.00e-01]
[2.30e+01 4.60e+00 3.60e+00 1.00e+00 2.00e-01]
[2.40e+01 5.10e+00 3.30e+00 1.70e+00 5.00e-01]
[2.50e+01 4.80e+00 3.40e+00 1.90e+00 2.00e-01]
[2.60e+01 5.00e+00 3.00e+00 1.60e+00 2.00e-01]
[2.70e+01 5.00e+00 3.40e+00 1.60e+00 4.00e-01]
[2.80e+01 5.20e+00 3.50e+00 1.50e+00 2.00e-01]
[2.90e+01 5.20e+00 3.40e+00 1.40e+00 2.00e-01]
[3.00e+01 4.70e+00 3.20e+00 1.60e+00 2.00e-01]
[3.10e+01 4.80e+00 3.10e+00 1.60e+00 2.00e-01]
[3.20e+01 5.40e+00 3.40e+00 1.50e+00 4.00e-01]
[3.30e+01 5.20e+00 4.10e+00 1.50e+00 1.00e-01]
[3.40e+01 5.50e+00 4.20e+00 1.40e+00 2.00e-01]
[3.50e+01 4.90e+00 3.10e+00 1.50e+00 1.00e-01]
[3.60e+01 5.00e+00 3.20e+00 1.20e+00 2.00e-01]
[3.70e+01 5.50e+00 3.50e+00 1.30e+00 2.00e-01]
[3.80e+01 4.90e+00 3.10e+00 1.50e+00 1.00e-01]
[3.90e+01 4.40e+00 3.00e+00 1.30e+00 2.00e-01]
[4.00e+01 5.10e+00 3.40e+00 1.50e+00 2.00e-01]
[4.10e+01 5.00e+00 3.50e+00 1.30e+00 3.00e-01]
[4.20e+01 4.50e+00 2.30e+00 1.30e+00 3.00e-01]
[4.30e+01 4.40e+00 3.20e+00 1.30e+00 2.00e-01]
[4.40e+01 5.00e+00 3.50e+00 1.60e+00 6.00e-01]
[4.50e+01 5.10e+00 3.80e+00 1.90e+00 4.00e-01]
[4.60e+01 4.80e+00 3.00e+00 1.40e+00 3.00e-01]
[4.70e+01 5.10e+00 3.80e+00 1.60e+00 2.00e-01]
[4.80e+01 4.60e+00 3.20e+00 1.40e+00 2.00e-01]
[4.90e+01 5.30e+00 3.70e+00 1.50e+00 2.00e-01]
[5.00e+01 5.00e+00 3.30e+00 1.40e+00 2.00e-01]
[5.10e+01 7.00e+00 3.20e+00 4.70e+00 1.40e+00]
[5.20e+01 6.40e+00 3.20e+00 4.50e+00 1.50e+00]
[5.30e+01 6.90e+00 3.10e+00 4.90e+00 1.50e+00]
[5.40e+01 5.50e+00 2.30e+00 4.00e+00 1.30e+00]
[5.50e+01 6.50e+00 2.80e+00 4.60e+00 1.50e+00]
[5.60e+01 5.70e+00 2.80e+00 4.50e+00 1.30e+00]
[5.70e+01 6.30e+00 3.30e+00 4.70e+00 1.60e+00]
[5.80e+01 4.90e+00 2.40e+00 3.30e+00 1.00e+00]
[5.90e+01 6.60e+00 2.90e+00 4.60e+00 1.30e+00]
[6.00e+01 5.20e+00 2.70e+00 3.90e+00 1.40e+00]
[6.10e+01 5.00e+00 2.00e+00 3.50e+00 1.00e+00]
[6.20e+01 5.90e+00 3.00e+00 4.20e+00 1.50e+00]
[6.30e+01 6.00e+00 2.20e+00 4.00e+00 1.00e+00]
[6.40e+01 6.10e+00 2.90e+00 4.70e+00 1.40e+00]
[6.50e+01 5.60e+00 2.90e+00 3.60e+00 1.30e+00]
[6.60e+01 6.70e+00 3.10e+00 4.40e+00 1.40e+00]
[6.70e+01 5.60e+00 3.00e+00 4.50e+00 1.50e+00]
[6.80e+01 5.80e+00 2.70e+00 4.10e+00 1.00e+00]
[6.90e+01 6.20e+00 2.20e+00 4.50e+00 1.50e+00]
[7.00e+01 5.60e+00 2.50e+00 3.90e+00 1.10e+00]
[7.10e+01 5.90e+00 3.20e+00 4.80e+00 1.80e+00]
[7.20e+01 6.10e+00 2.80e+00 4.00e+00 1.30e+00]
[7.30e+01 6.30e+00 2.50e+00 4.90e+00 1.50e+00]
[7.40e+01 6.10e+00 2.80e+00 4.70e+00 1.20e+00]
[7.50e+01 6.40e+00 2.90e+00 4.30e+00 1.30e+00]
[7.60e+01 6.60e+00 3.00e+00 4.40e+00 1.40e+00]
[7.70e+01 6.80e+00 2.80e+00 4.80e+00 1.40e+00]
[7.80e+01 6.70e+00 3.00e+00 5.00e+00 1.70e+00]
[7.90e+01 6.00e+00 2.90e+00 4.50e+00 1.50e+00]
[8.00e+01 5.70e+00 2.60e+00 3.50e+00 1.00e+00]
[8.10e+01 5.50e+00 2.40e+00 3.80e+00 1.10e+00]
[8.20e+01 5.50e+00 2.40e+00 3.70e+00 1.00e+00]
[8.30e+01 5.80e+00 2.70e+00 3.90e+00 1.20e+00]
[8.40e+01 6.00e+00 2.70e+00 5.10e+00 1.60e+00]
[8.50e+01 5.40e+00 3.00e+00 4.50e+00 1.50e+00]
[8.60e+01 6.00e+00 3.40e+00 4.50e+00 1.60e+00]
[8.70e+01 6.70e+00 3.10e+00 4.70e+00 1.50e+00]
[8.80e+01 6.30e+00 2.30e+00 4.40e+00 1.30e+00]
[8.90e+01 5.60e+00 3.00e+00 4.10e+00 1.30e+00]
[9.00e+01 5.50e+00 2.50e+00 4.00e+00 1.30e+00]
[9.10e+01 5.50e+00 2.60e+00 4.40e+00 1.20e+00]
[9.20e+01 6.10e+00 3.00e+00 4.60e+00 1.40e+00]
[9.30e+01 5.80e+00 2.60e+00 4.00e+00 1.20e+00]
[9.40e+01 5.00e+00 2.30e+00 3.30e+00 1.00e+00]
[9.50e+01 5.60e+00 2.70e+00 4.20e+00 1.30e+00]
[9.60e+01 5.70e+00 3.00e+00 4.20e+00 1.20e+00]
[9.70e+01 5.70e+00 2.90e+00 4.20e+00 1.30e+00]
[9.80e+01 6.20e+00 2.90e+00 4.30e+00 1.30e+00]
[9.90e+01 5.10e+00 2.50e+00 3.00e+00 1.10e+00]
[1.00e+02 5.70e+00 2.80e+00 4.10e+00 1.30e+00]
[1.01e+02 6.30e+00 3.30e+00 6.00e+00 2.50e+00]
[1.02e+02 5.80e+00 2.70e+00 5.10e+00 1.90e+00]
[1.03e+02 7.10e+00 3.00e+00 5.90e+00 2.10e+00]
[1.04e+02 6.30e+00 2.90e+00 5.60e+00 1.80e+00]
[1.05e+02 6.50e+00 3.00e+00 5.80e+00 2.20e+00]
[1.06e+02 7.60e+00 3.00e+00 6.60e+00 2.10e+00]
[1.07e+02 4.90e+00 2.50e+00 4.50e+00 1.70e+00]
[1.08e+02 7.30e+00 2.90e+00 6.30e+00 1.80e+00]
[1.09e+02 6.70e+00 2.50e+00 5.80e+00 1.80e+00]
[1.10e+02 7.20e+00 3.60e+00 6.10e+00 2.50e+00]
[1.11e+02 6.50e+00 3.20e+00 5.10e+00 2.00e+00]
[1.12e+02 6.40e+00 2.70e+00 5.30e+00 1.90e+00]
[1.13e+02 6.80e+00 3.00e+00 5.50e+00 2.10e+00]
[1.14e+02 5.70e+00 2.50e+00 5.00e+00 2.00e+00]
[1.15e+02 5.80e+00 2.80e+00 5.10e+00 2.40e+00]
[1.16e+02 6.40e+00 3.20e+00 5.30e+00 2.30e+00]
[1.17e+02 6.50e+00 3.00e+00 5.50e+00 1.80e+00]
[1.18e+02 7.70e+00 3.80e+00 6.70e+00 2.20e+00]
[1.19e+02 7.70e+00 2.60e+00 6.90e+00 2.30e+00]
[1.20e+02 6.00e+00 2.20e+00 5.00e+00 1.50e+00]
[1.21e+02 6.90e+00 3.20e+00 5.70e+00 2.30e+00]
[1.22e+02 5.60e+00 2.80e+00 4.90e+00 2.00e+00]
[1.23e+02 7.70e+00 2.80e+00 6.70e+00 2.00e+00]
[1.24e+02 6.30e+00 2.70e+00 4.90e+00 1.80e+00]
[1.25e+02 6.70e+00 3.30e+00 5.70e+00 2.10e+00]
[1.26e+02 7.20e+00 3.20e+00 6.00e+00 1.80e+00]
[1.27e+02 6.20e+00 2.80e+00 4.80e+00 1.80e+00]
[1.28e+02 6.10e+00 3.00e+00 4.90e+00 1.80e+00]
[1.29e+02 6.40e+00 2.80e+00 5.60e+00 2.10e+00]
[1.30e+02 7.20e+00 3.00e+00 5.80e+00 1.60e+00]
[1.31e+02 7.40e+00 2.80e+00 6.10e+00 1.90e+00]
[1.32e+02 7.90e+00 3.80e+00 6.40e+00 2.00e+00]
[1.33e+02 6.40e+00 2.80e+00 5.60e+00 2.20e+00]
[1.34e+02 6.30e+00 2.80e+00 5.10e+00 1.50e+00]
[1.35e+02 6.10e+00 2.60e+00 5.60e+00 1.40e+00]
[1.36e+02 7.70e+00 3.00e+00 6.10e+00 2.30e+00]
[1.37e+02 6.30e+00 3.40e+00 5.60e+00 2.40e+00]
[1.38e+02 6.40e+00 3.10e+00 5.50e+00 1.80e+00]
[1.39e+02 6.00e+00 3.00e+00 4.80e+00 1.80e+00]
[1.40e+02 6.90e+00 3.10e+00 5.40e+00 2.10e+00]
[1.41e+02 6.70e+00 3.10e+00 5.60e+00 2.40e+00]
[1.42e+02 6.90e+00 3.10e+00 5.10e+00 2.30e+00]
[1.43e+02 5.80e+00 2.70e+00 5.10e+00 1.90e+00]
[1.44e+02 6.80e+00 3.20e+00 5.90e+00 2.30e+00]
[1.45e+02 6.70e+00 3.30e+00 5.70e+00 2.50e+00]
[1.46e+02 6.70e+00 3.00e+00 5.20e+00 2.30e+00]
[1.47e+02 6.30e+00 2.50e+00 5.00e+00 1.90e+00]
[1.48e+02 6.50e+00 3.00e+00 5.20e+00 2.00e+00]
[1.49e+02 6.20e+00 3.40e+00 5.40e+00 2.30e+00]
[1.50e+02 5.90e+00 3.00e+00 5.10e+00 1.80e+00]] [[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[1 0 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 1 0 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]
[0 0 1 0]]
[[3. 1. 0.]
[2. 0. 1.]
[4. 0. 1.]
[3. 0. 1.]] [127500 106000 178100 140000]
CRIM ZN INDUS CHAS NOX ... DIS RAD TAX PTRATIO LSTAT
0 0.00632 18.0 2.31 0 0.538 ... 4.0900 1 296 15.3 4.98
1 0.02731 0.0 7.07 0 0.469 ... 4.9671 2 242 17.8 9.14
2 0.02729 0.0 7.07 0 0.469 ... 4.9671 2 242 17.8 4.03
3 0.03237 0.0 2.18 0 0.458 ... 6.0622 3 222 18.7 2.94
4 0.06905 0.0 2.18 0 0.458 ... 6.0622 3 222 18.7 5.33
.. ... ... ... ... ... ... ... ... ... ... ...
501 0.06263 0.0 11.93 0 0.573 ... 2.4786 1 273 21.0 9.67
502 0.04527 0.0 11.93 0 0.573 ... 2.2875 1 273 21.0 9.08
503 0.06076 0.0 11.93 0 0.573 ... 2.1675 1 273 21.0 5.64
504 0.10959 0.0 11.93 0 0.573 ... 2.3889 1 273 21.0 6.48
505 0.04741 0.0 11.93 0 0.573 ... 2.5050 1 273 21.0 7.88
[506 rows x 12 columns] 0 24.0
1 21.6
2 34.7
3 33.4
4 36.2
...
501 22.4
502 20.6
503 23.9
504 22.0
505 11.9
Name: MEDV, Length: 506, dtype: float64
Process finished with exit code 0
'''
- 转换为张量格式
import pandas as pd
import torch
#读取数据集
data_iris = pd.read_csv('Iris.csv')
data_house_tiny = pd.read_csv('house_tiny.csv')
data_boston_house_prices = pd.read_csv('boston_house_prices.csv')
#处理缺失值和离散值
input,output = data_iris.iloc[:,0:5],data_iris.iloc[:,5]
#使用均值处理缺失值
input = input.fillna(input.mean())
input = np.array(input)
#处理离散值
output = pd.get_dummies(output,dummy_na=True)
output = np.array(output)
#转换为张量形式
input_tensor,output_tensor = torch.tensor(input),torch.tensor(output)
print(input_tensor,output_tensor)
#读取输入输出
input_house_tiny,output_house_tiny = data_house_tiny.iloc[:,0:2],data_house_tiny.iloc[:,2]
#处理缺失值
input_house_tiny = input_house_tiny.fillna(input_house_tiny.mean())
#处理离散值
input_house_tiny = pd.get_dummies(input_house_tiny,dummy_na=True)
input_house_tiny,output_house_tiny = np.array(input_house_tiny),np.array(output_house_tiny)
#转换为张量形式
input_house_tiny_tensor,output_house_tiny_tensor = torch.tensor(input_house_tiny),torch.tensor(output_house_tiny)
print(input_house_tiny_tensor,output_house_tiny_tensor)
#读取输出
input_boston_house_prices,output_boston_house_prices = data_boston_house_prices.iloc[:,0:12],data_boston_house_prices.iloc[:,12]
#处理缺失值
input_boston_house_prices,output_boston_house_prices = input_boston_house_prices.fillna(input_boston_house_prices.mean()),output_boston_house_prices.fillna(output_boston_house_prices.mean())
#转换为张量形式
input_boston_house_prices_numpy,output_boston_house_prices_numpy = np.array(input_boston_house_prices),np.array(output_boston_house_prices)
input_boston_house_prices_tensor,output_boston_house_prices_tensor = torch.tensor(input_boston_house_prices_numpy),torch.tensor(output_boston_house_prices_numpy)
print(input_boston_house_prices_tensor,output_boston_house_prices_tensor)
'''
tensor([[1.0000e+00, 5.1000e+00, 3.5000e+00, 1.4000e+00, 2.0000e-01],
[2.0000e+00, 4.9000e+00, 3.0000e+00, 1.4000e+00, 2.0000e-01],
[3.0000e+00, 4.7000e+00, 3.2000e+00, 1.3000e+00, 2.0000e-01],
[4.0000e+00, 4.6000e+00, 3.1000e+00, 1.5000e+00, 2.0000e-01],
[5.0000e+00, 5.0000e+00, 3.6000e+00, 1.4000e+00, 2.0000e-01],
[6.0000e+00, 5.4000e+00, 3.9000e+00, 1.7000e+00, 4.0000e-01],
[7.0000e+00, 4.6000e+00, 3.4000e+00, 1.4000e+00, 3.0000e-01],
[8.0000e+00, 5.0000e+00, 3.4000e+00, 1.5000e+00, 2.0000e-01],
[9.0000e+00, 4.4000e+00, 2.9000e+00, 1.4000e+00, 2.0000e-01],
[1.0000e+01, 4.9000e+00, 3.1000e+00, 1.5000e+00, 1.0000e-01],
[1.1000e+01, 5.4000e+00, 3.7000e+00, 1.5000e+00, 2.0000e-01],
[1.2000e+01, 4.8000e+00, 3.4000e+00, 1.6000e+00, 2.0000e-01],
[1.3000e+01, 4.8000e+00, 3.0000e+00, 1.4000e+00, 1.0000e-01],
[1.4000e+01, 4.3000e+00, 3.0000e+00, 1.1000e+00, 1.0000e-01],
[1.5000e+01, 5.8000e+00, 4.0000e+00, 1.2000e+00, 2.0000e-01],
[1.6000e+01, 5.7000e+00, 4.4000e+00, 1.5000e+00, 4.0000e-01],
[1.7000e+01, 5.4000e+00, 3.9000e+00, 1.3000e+00, 4.0000e-01],
[1.8000e+01, 5.1000e+00, 3.5000e+00, 1.4000e+00, 3.0000e-01],
[1.9000e+01, 5.7000e+00, 3.8000e+00, 1.7000e+00, 3.0000e-01],
[2.0000e+01, 5.1000e+00, 3.8000e+00, 1.5000e+00, 3.0000e-01],
[2.1000e+01, 5.4000e+00, 3.4000e+00, 1.7000e+00, 2.0000e-01],
[2.2000e+01, 5.1000e+00, 3.7000e+00, 1.5000e+00, 4.0000e-01],
[2.3000e+01, 4.6000e+00, 3.6000e+00, 1.0000e+00, 2.0000e-01],
[2.4000e+01, 5.1000e+00, 3.3000e+00, 1.7000e+00, 5.0000e-01],
[2.5000e+01, 4.8000e+00, 3.4000e+00, 1.9000e+00, 2.0000e-01],
[2.6000e+01, 5.0000e+00, 3.0000e+00, 1.6000e+00, 2.0000e-01],
[2.7000e+01, 5.0000e+00, 3.4000e+00, 1.6000e+00, 4.0000e-01],
[2.8000e+01, 5.2000e+00, 3.5000e+00, 1.5000e+00, 2.0000e-01],
[2.9000e+01, 5.2000e+00, 3.4000e+00, 1.4000e+00, 2.0000e-01],
[3.0000e+01, 4.7000e+00, 3.2000e+00, 1.6000e+00, 2.0000e-01],
[3.1000e+01, 4.8000e+00, 3.1000e+00, 1.6000e+00, 2.0000e-01],
[3.2000e+01, 5.4000e+00, 3.4000e+00, 1.5000e+00, 4.0000e-01],
[3.3000e+01, 5.2000e+00, 4.1000e+00, 1.5000e+00, 1.0000e-01],
[3.4000e+01, 5.5000e+00, 4.2000e+00, 1.4000e+00, 2.0000e-01],
[3.5000e+01, 4.9000e+00, 3.1000e+00, 1.5000e+00, 1.0000e-01],
[3.6000e+01, 5.0000e+00, 3.2000e+00, 1.2000e+00, 2.0000e-01],
[3.7000e+01, 5.5000e+00, 3.5000e+00, 1.3000e+00, 2.0000e-01],
[3.8000e+01, 4.9000e+00, 3.1000e+00, 1.5000e+00, 1.0000e-01],
[3.9000e+01, 4.4000e+00, 3.0000e+00, 1.3000e+00, 2.0000e-01],
[4.0000e+01, 5.1000e+00, 3.4000e+00, 1.5000e+00, 2.0000e-01],
[4.1000e+01, 5.0000e+00, 3.5000e+00, 1.3000e+00, 3.0000e-01],
[4.2000e+01, 4.5000e+00, 2.3000e+00, 1.3000e+00, 3.0000e-01],
[4.3000e+01, 4.4000e+00, 3.2000e+00, 1.3000e+00, 2.0000e-01],
[4.4000e+01, 5.0000e+00, 3.5000e+00, 1.6000e+00, 6.0000e-01],
[4.5000e+01, 5.1000e+00, 3.8000e+00, 1.9000e+00, 4.0000e-01],
[4.6000e+01, 4.8000e+00, 3.0000e+00, 1.4000e+00, 3.0000e-01],
[4.7000e+01, 5.1000e+00, 3.8000e+00, 1.6000e+00, 2.0000e-01],
[4.8000e+01, 4.6000e+00, 3.2000e+00, 1.4000e+00, 2.0000e-01],
[4.9000e+01, 5.3000e+00, 3.7000e+00, 1.5000e+00, 2.0000e-01],
[5.0000e+01, 5.0000e+00, 3.3000e+00, 1.4000e+00, 2.0000e-01],
[5.1000e+01, 7.0000e+00, 3.2000e+00, 4.7000e+00, 1.4000e+00],
[5.2000e+01, 6.4000e+00, 3.2000e+00, 4.5000e+00, 1.5000e+00],
[5.3000e+01, 6.9000e+00, 3.1000e+00, 4.9000e+00, 1.5000e+00],
[5.4000e+01, 5.5000e+00, 2.3000e+00, 4.0000e+00, 1.3000e+00],
[5.5000e+01, 6.5000e+00, 2.8000e+00, 4.6000e+00, 1.5000e+00],
[5.6000e+01, 5.7000e+00, 2.8000e+00, 4.5000e+00, 1.3000e+00],
[5.7000e+01, 6.3000e+00, 3.3000e+00, 4.7000e+00, 1.6000e+00],
[5.8000e+01, 4.9000e+00, 2.4000e+00, 3.3000e+00, 1.0000e+00],
[5.9000e+01, 6.6000e+00, 2.9000e+00, 4.6000e+00, 1.3000e+00],
[6.0000e+01, 5.2000e+00, 2.7000e+00, 3.9000e+00, 1.4000e+00],
[6.1000e+01, 5.0000e+00, 2.0000e+00, 3.5000e+00, 1.0000e+00],
[6.2000e+01, 5.9000e+00, 3.0000e+00, 4.2000e+00, 1.5000e+00],
[6.3000e+01, 6.0000e+00, 2.2000e+00, 4.0000e+00, 1.0000e+00],
[6.4000e+01, 6.1000e+00, 2.9000e+00, 4.7000e+00, 1.4000e+00],
[6.5000e+01, 5.6000e+00, 2.9000e+00, 3.6000e+00, 1.3000e+00],
[6.6000e+01, 6.7000e+00, 3.1000e+00, 4.4000e+00, 1.4000e+00],
[6.7000e+01, 5.6000e+00, 3.0000e+00, 4.5000e+00, 1.5000e+00],
[6.8000e+01, 5.8000e+00, 2.7000e+00, 4.1000e+00, 1.0000e+00],
[6.9000e+01, 6.2000e+00, 2.2000e+00, 4.5000e+00, 1.5000e+00],
[7.0000e+01, 5.6000e+00, 2.5000e+00, 3.9000e+00, 1.1000e+00],
[7.1000e+01, 5.9000e+00, 3.2000e+00, 4.8000e+00, 1.8000e+00],
[7.2000e+01, 6.1000e+00, 2.8000e+00, 4.0000e+00, 1.3000e+00],
[7.3000e+01, 6.3000e+00, 2.5000e+00, 4.9000e+00, 1.5000e+00],
[7.4000e+01, 6.1000e+00, 2.8000e+00, 4.7000e+00, 1.2000e+00],
[7.5000e+01, 6.4000e+00, 2.9000e+00, 4.3000e+00, 1.3000e+00],
[7.6000e+01, 6.6000e+00, 3.0000e+00, 4.4000e+00, 1.4000e+00],
[7.7000e+01, 6.8000e+00, 2.8000e+00, 4.8000e+00, 1.4000e+00],
[7.8000e+01, 6.7000e+00, 3.0000e+00, 5.0000e+00, 1.7000e+00],
[7.9000e+01, 6.0000e+00, 2.9000e+00, 4.5000e+00, 1.5000e+00],
[8.0000e+01, 5.7000e+00, 2.6000e+00, 3.5000e+00, 1.0000e+00],
[8.1000e+01, 5.5000e+00, 2.4000e+00, 3.8000e+00, 1.1000e+00],
[8.2000e+01, 5.5000e+00, 2.4000e+00, 3.7000e+00, 1.0000e+00],
[8.3000e+01, 5.8000e+00, 2.7000e+00, 3.9000e+00, 1.2000e+00],
[8.4000e+01, 6.0000e+00, 2.7000e+00, 5.1000e+00, 1.6000e+00],
[8.5000e+01, 5.4000e+00, 3.0000e+00, 4.5000e+00, 1.5000e+00],
[8.6000e+01, 6.0000e+00, 3.4000e+00, 4.5000e+00, 1.6000e+00],
[8.7000e+01, 6.7000e+00, 3.1000e+00, 4.7000e+00, 1.5000e+00],
[8.8000e+01, 6.3000e+00, 2.3000e+00, 4.4000e+00, 1.3000e+00],
[8.9000e+01, 5.6000e+00, 3.0000e+00, 4.1000e+00, 1.3000e+00],
[9.0000e+01, 5.5000e+00, 2.5000e+00, 4.0000e+00, 1.3000e+00],
[9.1000e+01, 5.5000e+00, 2.6000e+00, 4.4000e+00, 1.2000e+00],
[9.2000e+01, 6.1000e+00, 3.0000e+00, 4.6000e+00, 1.4000e+00],
[9.3000e+01, 5.8000e+00, 2.6000e+00, 4.0000e+00, 1.2000e+00],
[9.4000e+01, 5.0000e+00, 2.3000e+00, 3.3000e+00, 1.0000e+00],
[9.5000e+01, 5.6000e+00, 2.7000e+00, 4.2000e+00, 1.3000e+00],
[9.6000e+01, 5.7000e+00, 3.0000e+00, 4.2000e+00, 1.2000e+00],
[9.7000e+01, 5.7000e+00, 2.9000e+00, 4.2000e+00, 1.3000e+00],
[9.8000e+01, 6.2000e+00, 2.9000e+00, 4.3000e+00, 1.3000e+00],
[9.9000e+01, 5.1000e+00, 2.5000e+00, 3.0000e+00, 1.1000e+00],
[1.0000e+02, 5.7000e+00, 2.8000e+00, 4.1000e+00, 1.3000e+00],
[1.0100e+02, 6.3000e+00, 3.3000e+00, 6.0000e+00, 2.5000e+00],
[1.0200e+02, 5.8000e+00, 2.7000e+00, 5.1000e+00, 1.9000e+00],
[1.0300e+02, 7.1000e+00, 3.0000e+00, 5.9000e+00, 2.1000e+00],
[1.0400e+02, 6.3000e+00, 2.9000e+00, 5.6000e+00, 1.8000e+00],
[1.0500e+02, 6.5000e+00, 3.0000e+00, 5.8000e+00, 2.2000e+00],
[1.0600e+02, 7.6000e+00, 3.0000e+00, 6.6000e+00, 2.1000e+00],
[1.0700e+02, 4.9000e+00, 2.5000e+00, 4.5000e+00, 1.7000e+00],
[1.0800e+02, 7.3000e+00, 2.9000e+00, 6.3000e+00, 1.8000e+00],
[1.0900e+02, 6.7000e+00, 2.5000e+00, 5.8000e+00, 1.8000e+00],
[1.1000e+02, 7.2000e+00, 3.6000e+00, 6.1000e+00, 2.5000e+00],
[1.1100e+02, 6.5000e+00, 3.2000e+00, 5.1000e+00, 2.0000e+00],
[1.1200e+02, 6.4000e+00, 2.7000e+00, 5.3000e+00, 1.9000e+00],
[1.1300e+02, 6.8000e+00, 3.0000e+00, 5.5000e+00, 2.1000e+00],
[1.1400e+02, 5.7000e+00, 2.5000e+00, 5.0000e+00, 2.0000e+00],
[1.1500e+02, 5.8000e+00, 2.8000e+00, 5.1000e+00, 2.4000e+00],
[1.1600e+02, 6.4000e+00, 3.2000e+00, 5.3000e+00, 2.3000e+00],
[1.1700e+02, 6.5000e+00, 3.0000e+00, 5.5000e+00, 1.8000e+00],
[1.1800e+02, 7.7000e+00, 3.8000e+00, 6.7000e+00, 2.2000e+00],
[1.1900e+02, 7.7000e+00, 2.6000e+00, 6.9000e+00, 2.3000e+00],
[1.2000e+02, 6.0000e+00, 2.2000e+00, 5.0000e+00, 1.5000e+00],
[1.2100e+02, 6.9000e+00, 3.2000e+00, 5.7000e+00, 2.3000e+00],
[1.2200e+02, 5.6000e+00, 2.8000e+00, 4.9000e+00, 2.0000e+00],
[1.2300e+02, 7.7000e+00, 2.8000e+00, 6.7000e+00, 2.0000e+00],
[1.2400e+02, 6.3000e+00, 2.7000e+00, 4.9000e+00, 1.8000e+00],
[1.2500e+02, 6.7000e+00, 3.3000e+00, 5.7000e+00, 2.1000e+00],
[1.2600e+02, 7.2000e+00, 3.2000e+00, 6.0000e+00, 1.8000e+00],
[1.2700e+02, 6.2000e+00, 2.8000e+00, 4.8000e+00, 1.8000e+00],
[1.2800e+02, 6.1000e+00, 3.0000e+00, 4.9000e+00, 1.8000e+00],
[1.2900e+02, 6.4000e+00, 2.8000e+00, 5.6000e+00, 2.1000e+00],
[1.3000e+02, 7.2000e+00, 3.0000e+00, 5.8000e+00, 1.6000e+00],
[1.3100e+02, 7.4000e+00, 2.8000e+00, 6.1000e+00, 1.9000e+00],
[1.3200e+02, 7.9000e+00, 3.8000e+00, 6.4000e+00, 2.0000e+00],
[1.3300e+02, 6.4000e+00, 2.8000e+00, 5.6000e+00, 2.2000e+00],
[1.3400e+02, 6.3000e+00, 2.8000e+00, 5.1000e+00, 1.5000e+00],
[1.3500e+02, 6.1000e+00, 2.6000e+00, 5.6000e+00, 1.4000e+00],
[1.3600e+02, 7.7000e+00, 3.0000e+00, 6.1000e+00, 2.3000e+00],
[1.3700e+02, 6.3000e+00, 3.4000e+00, 5.6000e+00, 2.4000e+00],
[1.3800e+02, 6.4000e+00, 3.1000e+00, 5.5000e+00, 1.8000e+00],
[1.3900e+02, 6.0000e+00, 3.0000e+00, 4.8000e+00, 1.8000e+00],
[1.4000e+02, 6.9000e+00, 3.1000e+00, 5.4000e+00, 2.1000e+00],
[1.4100e+02, 6.7000e+00, 3.1000e+00, 5.6000e+00, 2.4000e+00],
[1.4200e+02, 6.9000e+00, 3.1000e+00, 5.1000e+00, 2.3000e+00],
[1.4300e+02, 5.8000e+00, 2.7000e+00, 5.1000e+00, 1.9000e+00],
[1.4400e+02, 6.8000e+00, 3.2000e+00, 5.9000e+00, 2.3000e+00],
[1.4500e+02, 6.7000e+00, 3.3000e+00, 5.7000e+00, 2.5000e+00],
[1.4600e+02, 6.7000e+00, 3.0000e+00, 5.2000e+00, 2.3000e+00],
[1.4700e+02, 6.3000e+00, 2.5000e+00, 5.0000e+00, 1.9000e+00],
[1.4800e+02, 6.5000e+00, 3.0000e+00, 5.2000e+00, 2.0000e+00],
[1.4900e+02, 6.2000e+00, 3.4000e+00, 5.4000e+00, 2.3000e+00],
[1.5000e+02, 5.9000e+00, 3.0000e+00, 5.1000e+00, 1.8000e+00]],
dtype=torch.float64) tensor([[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[1, 0, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 1, 0, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0],
[0, 0, 1, 0]], dtype=torch.uint8)
tensor([[3., 1., 0.],
[2., 0., 1.],
[4., 0., 1.],
[3., 0., 1.]], dtype=torch.float64) tensor([127500, 106000, 178100, 140000])
tensor([[6.3200e-03, 1.8000e+01, 2.3100e+00, ..., 2.9600e+02, 1.5300e+01,
4.9800e+00],
[2.7310e-02, 0.0000e+00, 7.0700e+00, ..., 2.4200e+02, 1.7800e+01,
9.1400e+00],
[2.7290e-02, 0.0000e+00, 7.0700e+00, ..., 2.4200e+02, 1.7800e+01,
4.0300e+00],
...,
[6.0760e-02, 0.0000e+00, 1.1930e+01, ..., 2.7300e+02, 2.1000e+01,
5.6400e+00],
[1.0959e-01, 0.0000e+00, 1.1930e+01, ..., 2.7300e+02, 2.1000e+01,
6.4800e+00],
[4.7410e-02, 0.0000e+00, 1.1930e+01, ..., 2.7300e+02, 2.1000e+01,
7.8800e+00]], dtype=torch.float64) tensor([24.0000, 21.6000, 34.7000, 33.4000, 36.2000, 28.7000, 22.9000, 27.1000,
16.5000, 18.9000, 15.0000, 18.9000, 21.7000, 20.4000, 18.2000, 19.9000,
23.1000, 17.5000, 20.2000, 18.2000, 13.6000, 19.6000, 15.2000, 14.5000,
15.6000, 13.9000, 16.6000, 14.8000, 18.4000, 21.0000, 12.7000, 14.5000,
13.2000, 13.1000, 13.5000, 18.9000, 20.0000, 21.0000, 24.7000, 30.8000,
34.9000, 26.6000, 25.3000, 24.7000, 21.2000, 19.3000, 20.0000, 16.6000,
14.4000, 19.4000, 19.7000, 20.5000, 25.0000, 23.4000, 18.9000, 35.4000,
24.7000, 31.6000, 23.3000, 19.6000, 18.7000, 16.0000, 22.2000, 25.0000,
33.0000, 23.5000, 19.4000, 22.0000, 17.4000, 20.9000, 24.2000, 21.7000,
22.8000, 23.4000, 24.1000, 21.4000, 20.0000, 20.8000, 21.2000, 20.3000,
28.0000, 23.9000, 24.8000, 22.9000, 23.9000, 26.6000, 22.5000, 22.2000,
23.6000, 28.7000, 22.6000, 22.0000, 22.9000, 25.0000, 20.6000, 28.4000,
21.4000, 38.7000, 43.8000, 33.2000, 27.5000, 26.5000, 18.6000, 19.3000,
20.1000, 19.5000, 19.5000, 20.4000, 19.8000, 19.4000, 21.7000, 22.8000,
18.8000, 18.7000, 18.5000, 18.3000, 21.2000, 19.2000, 20.4000, 19.3000,
22.0000, 20.3000, 20.5000, 17.3000, 18.8000, 21.4000, 15.7000, 16.2000,
18.0000, 14.3000, 19.2000, 19.6000, 23.0000, 18.4000, 15.6000, 18.1000,
17.4000, 17.1000, 13.3000, 17.8000, 14.0000, 14.4000, 13.4000, 15.6000,
11.8000, 13.8000, 15.6000, 14.6000, 17.8000, 15.4000, 21.5000, 19.6000,
15.3000, 19.4000, 17.0000, 15.6000, 13.1000, 41.3000, 24.3000, 23.3000,
27.0000, 50.0000, 50.0000, 50.0000, 22.7000, 25.0000, 50.0000, 23.8000,
23.8000, 22.3000, 17.4000, 19.1000, 23.1000, 23.6000, 22.6000, 29.4000,
23.2000, 24.6000, 29.9000, 37.2000, 39.8000, 36.2000, 37.9000, 32.5000,
26.4000, 29.6000, 50.0000, 32.0000, 29.8000, 34.9000, 37.0000, 30.5000,
36.4000, 31.1000, 29.1000, 50.0000, 33.3000, 30.3000, 34.6000, 34.9000,
32.9000, 24.1000, 42.3000, 48.5000, 50.0000, 22.6000, 24.4000, 22.5000,
24.4000, 20.0000, 21.7000, 19.3000, 22.4000, 28.1000, 23.7000, 25.0000,
23.3000, 28.7000, 21.5000, 23.0000, 26.7000, 21.7000, 27.5000, 30.1000,
44.8000, 50.0000, 37.6000, 31.6000, 46.7000, 31.5000, 24.3000, 31.7000,
41.7000, 48.3000, 29.0000, 24.0000, 25.1000, 31.5000, 23.7000, 23.3000,
22.0000, 20.1000, 22.2000, 23.7000, 17.6000, 18.5000, 24.3000, 20.5000,
24.5000, 26.2000, 24.4000, 24.8000, 29.6000, 42.8000, 21.9000, 20.9000,
44.0000, 50.0000, 36.0000, 30.1000, 33.8000, 43.1000, 48.8000, 31.0000,
36.5000, 22.8000, 30.7000, 50.0000, 43.5000, 20.7000, 21.1000, 25.2000,
24.4000, 35.2000, 32.4000, 32.0000, 33.2000, 33.1000, 29.1000, 35.1000,
45.4000, 35.4000, 46.0000, 50.0000, 32.2000, 22.0000, 20.1000, 23.2000,
22.3000, 24.8000, 28.5000, 37.3000, 27.9000, 23.9000, 21.7000, 28.6000,
27.1000, 20.3000, 22.5000, 29.0000, 24.8000, 22.0000, 26.4000, 33.1000,
36.1000, 28.4000, 33.4000, 28.2000, 22.8000, 20.3000, 16.1000, 22.1000,
19.4000, 21.6000, 23.8000, 16.2000, 17.8000, 19.8000, 23.1000, 21.0000,
23.8000, 23.1000, 20.4000, 18.5000, 25.0000, 24.6000, 23.0000, 22.2000,
19.3000, 22.6000, 19.8000, 17.1000, 19.4000, 22.2000, 20.7000, 21.1000,
19.5000, 18.5000, 20.6000, 19.0000, 18.7000, 32.7000, 16.5000, 23.9000,
31.2000, 17.5000, 17.2000, 23.1000, 24.5000, 26.6000, 22.9000, 24.1000,
18.6000, 30.1000, 18.2000, 20.6000, 17.8000, 21.7000, 22.7000, 22.6000,
25.0000, 19.9000, 20.8000, 16.8000, 21.9000, 27.5000, 21.9000, 23.1000,
50.0000, 50.0000, 50.0000, 50.0000, 50.0000, 13.8000, 13.8000, 15.0000,
13.9000, 13.3000, 13.1000, 10.2000, 10.4000, 10.9000, 11.3000, 12.3000,
8.8000, 7.2000, 10.5000, 7.4000, 10.2000, 11.5000, 15.1000, 23.2000,
9.7000, 13.8000, 12.7000, 13.1000, 12.5000, 8.5000, 5.0000, 6.3000,
5.6000, 7.2000, 12.1000, 8.3000, 8.5000, 5.0000, 11.9000, 27.9000,
17.2000, 27.5000, 15.0000, 17.2000, 17.9000, 16.3000, 7.0000, 7.2000,
7.5000, 10.4000, 8.8000, 8.4000, 16.7000, 14.2000, 20.8000, 13.4000,
11.7000, 8.3000, 10.2000, 10.9000, 11.0000, 9.5000, 14.5000, 14.1000,
16.1000, 14.3000, 11.7000, 13.4000, 9.6000, 8.7000, 8.4000, 12.8000,
10.5000, 17.1000, 18.4000, 15.4000, 10.8000, 11.8000, 14.9000, 12.6000,
14.1000, 13.0000, 13.4000, 15.2000, 16.1000, 17.8000, 14.9000, 14.1000,
12.7000, 13.5000, 14.9000, 20.0000, 16.4000, 17.7000, 19.5000, 20.2000,
21.4000, 19.9000, 19.0000, 19.1000, 19.1000, 20.1000, 19.9000, 19.6000,
23.2000, 29.8000, 13.8000, 13.3000, 16.7000, 12.0000, 14.6000, 21.4000,
23.0000, 23.7000, 25.0000, 21.8000, 20.6000, 21.2000, 19.1000, 20.6000,
15.2000, 7.0000, 8.1000, 13.6000, 20.1000, 21.8000, 24.5000, 23.1000,
19.7000, 18.3000, 21.2000, 17.5000, 16.8000, 22.4000, 20.6000, 23.9000,
22.0000, 11.9000], dtype=torch.float64)
Process finished with exit code 0
'''
心得体会:这次的难度比上次略微有提高。
第一部分：回顾了一下张量和算子的概念，以及张量和numpy之间的区别和优缺点。
第二部分：回顾了一下基本的tensor处理数据的函数。函数太多，比较麻烦，不过天上没有白掉的馅饼，也算加深了印象。
第三部分: 是一个崭新的部分，之前机器学习老师给的是已经处理好的数据集，没有进行数据预处理这方面，这部分也算了解了一些，处理缺失值有两种办法，一种是插值，一种是删除，插值的化可以用均值插值，也可以用中位数进行插值。删除直接删除其他元素就好了，对于一些语义，不是用数字进行表示的，例如玫瑰花、西兰花，可以转换为数据种类长度的的维度的变量，然后用[1,0]这种表示，例如[1,0]表示的是玫瑰花。[0,1]表示的是西兰花。
在进行张量转换的时候，直接使用读出来的csv文件会报错，需要用numpy进行转换，在转换为tensor变量，numpy相当于一个不收钱的中间商。
传送门:河大最牛老师
