some part of this notebook are copied from Jose Marcial Portilla works.
Datasets with PyTorch
Perform standard imports
import torch
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline
Loading data from files
We’ve seen how to load NumPy arrays into PyTorch, and anyone familiar with pandas.read_csv() can use it to prepare data before forming tensors. Here we’ll load the iris flower dataset saved as a .csv file.
# download csv file from github
!wget https://raw.githubusercontent.com/alisharifi2000/LoopAcademyMedicalImageProcessing/main/pytorch/Data/IRIS.csv
--2023-02-18 08:40:49-- https://raw.githubusercontent.com/alisharifi2000/LoopAcademyMedicalImageProcessing/main/pytorch/Data/IRIS.csv
Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 185.199.109.133, 185.199.108.133, 185.199.110.133, ...
Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|185.199.109.133|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 4617 (4.5K) [text/plain]
Saving to: ‘IRIS.csv.1’
IRIS.csv.1 100%[===================>] 4.51K --.-KB/s in 0s
2023-02-18 08:40:49 (58.6 MB/s) - ‘IRIS.csv.1’ saved [4617/4617]
df = pd.read_csv('/content/IRIS.csv')
df.head()
| sepal_length | sepal_width | petal_length | petal_width | species | |
|---|---|---|---|---|---|
| 0 | 5.1 | 3.5 | 1.4 | 0.2 | Iris-setosa |
| 1 | 4.9 | 3.0 | 1.4 | 0.2 | Iris-setosa |
| 2 | 4.7 | 3.2 | 1.3 | 0.2 | Iris-setosa |
| 3 | 4.6 | 3.1 | 1.5 | 0.2 | Iris-setosa |
| 4 | 5.0 | 3.6 | 1.4 | 0.2 | Iris-setosa |
df.shape
(150, 5)
df.species.unique()
array(['Iris-setosa', 'Iris-versicolor', 'Iris-virginica'], dtype=object)
Plot the data
fig, axes = plt.subplots(nrows=2, ncols=2, figsize=(10,7))
fig.tight_layout()
plots = [(0,1),(2,3),(0,2),(1,3)]
colors = ['b', 'r', 'g']
labels = ['Iris-setosa','Iris-versicolor','Iris-virginica']
for i, ax in enumerate(axes.flat):
for j, label in enumerate(labels):
x = df.columns[plots[i][0]]
y = df.columns[plots[i][1]]
ax.scatter(df[df['species']== label][x], df[df['species']== label][y],
color=colors[j])
ax.set(xlabel=x, ylabel=y)
fig.legend(labels=labels, loc=3, bbox_to_anchor=(1.0,0.85))
plt.show()
The iris dataset consists of 50 samples each from three species of Iris (Iris setosa, Iris virginica and Iris versicolor), for 150 total samples. We have four features (sepal length & width, petal length & width) and three unique labels:
- Iris setosa
- Iris virginica
- Iris versicolor
The classic method for building train/test split tensors
Before introducing PyTorch’s Dataset and DataLoader classes, we’ll take a quick look at the alternative.
hint we should label encoding or one-hot encoding target. because labels must be numeric and not string!
we can use sklearn package but in this code we use pandas replace method.
repl = {'Iris-setosa':1, 'Iris-versicolor':2 , 'Iris-virginica':3}
df['species'] = df['species'].replace(repl)
df['species'].unique()
array([1, 2, 3])
from sklearn.model_selection import train_test_split
train_X, test_X, train_y, test_y = train_test_split(df.drop('species',axis=1).values,
df['species'].values, test_size=0.2,
random_state=33)
test_X.shape
(30, 4)
X_train = torch.tensor(train_X)
X_test = torch.tensor(test_X)
y_train = torch.tensor(train_y).reshape(-1, 1)
y_test = torch.tensor(test_y).reshape(-1, 1)
print(f'Training size: {len(y_train)}')
labels, counts = y_train.unique(return_counts=True)
print(f'Labels: {labels}\nCounts: {counts}')
Training size: 120
Labels: tensor([1, 2, 3])
Counts: tensor([42, 42, 36])
NOTE: The importance of a balanced training set is discussed in A systematic study of the class imbalance problem in convolutional neural networks by Mateusz Buda, Atsuto Maki, Maciej A. Mazurowski (10/15/17, latest rev 10/13/18) https://arxiv.org/abs/1710.05381
For example, the authors show that oversampling a less common class so that it matches the more common classes is always the preferred choice.
For example, the authors show that oversampling a less common class so that it matches the more common classes is always the preferred choice.
X_train.size()
torch.Size([120, 4])
y_train.size()
torch.Size([120, 1])
Using PyTorch’s Dataset and DataLoader classes
A far better alternative is to leverage PyTorch’s Dataset and DataLoader</strong></a> classes.
Usually, to set up a Dataset specific to our investigation we would define our own custom class that inherits from torch.utils.data.Dataset (we’ll do this in the CNN section). For now, we can use the built-in TensorDataset class.
from torch.utils.data import TensorDataset, DataLoader
data = df.drop('species',axis=1).values
labels = df['species'].values
iris = TensorDataset(torch.tensor(data),torch.tensor(labels))
len(iris)
150
type(iris)
torch.utils.data.dataset.TensorDataset
for i in iris:
print(i)
(tensor([5.1000, 3.5000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.9000, 3.0000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.7000, 3.2000, 1.3000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.6000, 3.1000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.6000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.4000, 3.9000, 1.7000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([4.6000, 3.4000, 1.4000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.4000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.4000, 2.9000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([5.4000, 3.7000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.8000, 3.4000, 1.6000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.8000, 3.0000, 1.4000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([4.3000, 3.0000, 1.1000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([5.8000, 4.0000, 1.2000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.7000, 4.4000, 1.5000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([5.4000, 3.9000, 1.3000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.5000, 1.4000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([5.7000, 3.8000, 1.7000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.8000, 1.5000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([5.4000, 3.4000, 1.7000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.7000, 1.5000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([4.6000, 3.6000, 1.0000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.3000, 1.7000, 0.5000], dtype=torch.float64), tensor(1))
(tensor([4.8000, 3.4000, 1.9000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.0000, 1.6000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.4000, 1.6000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([5.2000, 3.5000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.2000, 3.4000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.7000, 3.2000, 1.6000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.8000, 3.1000, 1.6000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.4000, 3.4000, 1.5000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([5.2000, 4.1000, 1.5000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([5.5000, 4.2000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.2000, 1.2000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.5000, 3.5000, 1.3000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor(1))
(tensor([4.4000, 3.0000, 1.3000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.4000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.5000, 1.3000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([4.5000, 2.3000, 1.3000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([4.4000, 3.2000, 1.3000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.5000, 1.6000, 0.6000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.8000, 1.9000, 0.4000], dtype=torch.float64), tensor(1))
(tensor([4.8000, 3.0000, 1.4000, 0.3000], dtype=torch.float64), tensor(1))
(tensor([5.1000, 3.8000, 1.6000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([4.6000, 3.2000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.3000, 3.7000, 1.5000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([5.0000, 3.3000, 1.4000, 0.2000], dtype=torch.float64), tensor(1))
(tensor([7.0000, 3.2000, 4.7000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([6.4000, 3.2000, 4.5000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([6.9000, 3.1000, 4.9000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([5.5000, 2.3000, 4.0000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.5000, 2.8000, 4.6000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([5.7000, 2.8000, 4.5000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.3000, 3.3000, 4.7000, 1.6000], dtype=torch.float64), tensor(2))
(tensor([4.9000, 2.4000, 3.3000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([6.6000, 2.9000, 4.6000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.2000, 2.7000, 3.9000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([5.0000, 2.0000, 3.5000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([5.9000, 3.0000, 4.2000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([6.0000, 2.2000, 4.0000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([6.1000, 2.9000, 4.7000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([5.6000, 2.9000, 3.6000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.7000, 3.1000, 4.4000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([5.6000, 3.0000, 4.5000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([5.8000, 2.7000, 4.1000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([6.2000, 2.2000, 4.5000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([5.6000, 2.5000, 3.9000, 1.1000], dtype=torch.float64), tensor(2))
(tensor([5.9000, 3.2000, 4.8000, 1.8000], dtype=torch.float64), tensor(2))
(tensor([6.1000, 2.8000, 4.0000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.3000, 2.5000, 4.9000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([6.1000, 2.8000, 4.7000, 1.2000], dtype=torch.float64), tensor(2))
(tensor([6.4000, 2.9000, 4.3000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.6000, 3.0000, 4.4000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([6.8000, 2.8000, 4.8000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([6.7000, 3.0000, 5.0000, 1.7000], dtype=torch.float64), tensor(2))
(tensor([6.0000, 2.9000, 4.5000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([5.7000, 2.6000, 3.5000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([5.5000, 2.4000, 3.8000, 1.1000], dtype=torch.float64), tensor(2))
(tensor([5.5000, 2.4000, 3.7000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([5.8000, 2.7000, 3.9000, 1.2000], dtype=torch.float64), tensor(2))
(tensor([6.0000, 2.7000, 5.1000, 1.6000], dtype=torch.float64), tensor(2))
(tensor([5.4000, 3.0000, 4.5000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([6.0000, 3.4000, 4.5000, 1.6000], dtype=torch.float64), tensor(2))
(tensor([6.7000, 3.1000, 4.7000, 1.5000], dtype=torch.float64), tensor(2))
(tensor([6.3000, 2.3000, 4.4000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.6000, 3.0000, 4.1000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.5000, 2.5000, 4.0000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.5000, 2.6000, 4.4000, 1.2000], dtype=torch.float64), tensor(2))
(tensor([6.1000, 3.0000, 4.6000, 1.4000], dtype=torch.float64), tensor(2))
(tensor([5.8000, 2.6000, 4.0000, 1.2000], dtype=torch.float64), tensor(2))
(tensor([5.0000, 2.3000, 3.3000, 1.0000], dtype=torch.float64), tensor(2))
(tensor([5.6000, 2.7000, 4.2000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.7000, 3.0000, 4.2000, 1.2000], dtype=torch.float64), tensor(2))
(tensor([5.7000, 2.9000, 4.2000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.2000, 2.9000, 4.3000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([5.1000, 2.5000, 3.0000, 1.1000], dtype=torch.float64), tensor(2))
(tensor([5.7000, 2.8000, 4.1000, 1.3000], dtype=torch.float64), tensor(2))
(tensor([6.3000, 3.3000, 6.0000, 2.5000], dtype=torch.float64), tensor(3))
(tensor([5.8000, 2.7000, 5.1000, 1.9000], dtype=torch.float64), tensor(3))
(tensor([7.1000, 3.0000, 5.9000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([6.3000, 2.9000, 5.6000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.5000, 3.0000, 5.8000, 2.2000], dtype=torch.float64), tensor(3))
(tensor([7.6000, 3.0000, 6.6000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([4.9000, 2.5000, 4.5000, 1.7000], dtype=torch.float64), tensor(3))
(tensor([7.3000, 2.9000, 6.3000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.7000, 2.5000, 5.8000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([7.2000, 3.6000, 6.1000, 2.5000], dtype=torch.float64), tensor(3))
(tensor([6.5000, 3.2000, 5.1000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([6.4000, 2.7000, 5.3000, 1.9000], dtype=torch.float64), tensor(3))
(tensor([6.8000, 3.0000, 5.5000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([5.7000, 2.5000, 5.0000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([5.8000, 2.8000, 5.1000, 2.4000], dtype=torch.float64), tensor(3))
(tensor([6.4000, 3.2000, 5.3000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([6.5000, 3.0000, 5.5000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([7.7000, 3.8000, 6.7000, 2.2000], dtype=torch.float64), tensor(3))
(tensor([7.7000, 2.6000, 6.9000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([6.0000, 2.2000, 5.0000, 1.5000], dtype=torch.float64), tensor(3))
(tensor([6.9000, 3.2000, 5.7000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([5.6000, 2.8000, 4.9000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([7.7000, 2.8000, 6.7000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([6.3000, 2.7000, 4.9000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.7000, 3.3000, 5.7000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([7.2000, 3.2000, 6.0000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.2000, 2.8000, 4.8000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.1000, 3.0000, 4.9000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.4000, 2.8000, 5.6000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([7.2000, 3.0000, 5.8000, 1.6000], dtype=torch.float64), tensor(3))
(tensor([7.4000, 2.8000, 6.1000, 1.9000], dtype=torch.float64), tensor(3))
(tensor([7.9000, 3.8000, 6.4000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([6.4000, 2.8000, 5.6000, 2.2000], dtype=torch.float64), tensor(3))
(tensor([6.3000, 2.8000, 5.1000, 1.5000], dtype=torch.float64), tensor(3))
(tensor([6.1000, 2.6000, 5.6000, 1.4000], dtype=torch.float64), tensor(3))
(tensor([7.7000, 3.0000, 6.1000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([6.3000, 3.4000, 5.6000, 2.4000], dtype=torch.float64), tensor(3))
(tensor([6.4000, 3.1000, 5.5000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.0000, 3.0000, 4.8000, 1.8000], dtype=torch.float64), tensor(3))
(tensor([6.9000, 3.1000, 5.4000, 2.1000], dtype=torch.float64), tensor(3))
(tensor([6.7000, 3.1000, 5.6000, 2.4000], dtype=torch.float64), tensor(3))
(tensor([6.9000, 3.1000, 5.1000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([5.8000, 2.7000, 5.1000, 1.9000], dtype=torch.float64), tensor(3))
(tensor([6.8000, 3.2000, 5.9000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([6.7000, 3.3000, 5.7000, 2.5000], dtype=torch.float64), tensor(3))
(tensor([6.7000, 3.0000, 5.2000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([6.3000, 2.5000, 5.0000, 1.9000], dtype=torch.float64), tensor(3))
(tensor([6.5000, 3.0000, 5.2000, 2.0000], dtype=torch.float64), tensor(3))
(tensor([6.2000, 3.4000, 5.4000, 2.3000], dtype=torch.float64), tensor(3))
(tensor([5.9000, 3.0000, 5.1000, 1.8000], dtype=torch.float64), tensor(3))
Once we have a dataset we can wrap it with a DataLoader. This gives us a powerful sampler that provides single- or multi-process iterators over the dataset.
iris_loader = DataLoader(iris, batch_size=15, shuffle=True)
print(iris_loader)
<torch.utils.data.dataloader.DataLoader object at 0x7fad2ac187f0>
for i_batch, sample_batched in enumerate(iris_loader):
print(i_batch, sample_batched)
0 [tensor([[5.6000, 2.5000, 3.9000, 1.1000],
[6.8000, 3.2000, 5.9000, 2.3000],
[5.1000, 3.5000, 1.4000, 0.2000],
[7.2000, 3.0000, 5.8000, 1.6000],
[6.3000, 2.5000, 5.0000, 1.9000],
[6.0000, 3.0000, 4.8000, 1.8000],
[6.4000, 2.8000, 5.6000, 2.2000],
[5.4000, 3.9000, 1.7000, 0.4000],
[5.5000, 2.3000, 4.0000, 1.3000],
[6.5000, 2.8000, 4.6000, 1.5000],
[5.7000, 2.8000, 4.1000, 1.3000],
[6.1000, 3.0000, 4.9000, 1.8000],
[5.7000, 2.9000, 4.2000, 1.3000],
[4.4000, 3.0000, 1.3000, 0.2000],
[6.3000, 3.4000, 5.6000, 2.4000]], dtype=torch.float64), tensor([2, 3, 1, 3, 3, 3, 3, 1, 2, 2, 2, 3, 2, 1, 3])]
1 [tensor([[6.1000, 2.9000, 4.7000, 1.4000],
[6.0000, 2.9000, 4.5000, 1.5000],
[5.1000, 3.8000, 1.9000, 0.4000],
[7.0000, 3.2000, 4.7000, 1.4000],
[5.6000, 2.7000, 4.2000, 1.3000],
[6.5000, 3.0000, 5.8000, 2.2000],
[5.6000, 3.0000, 4.5000, 1.5000],
[5.5000, 2.6000, 4.4000, 1.2000],
[6.9000, 3.1000, 5.1000, 2.3000],
[5.5000, 2.4000, 3.8000, 1.1000],
[5.4000, 3.7000, 1.5000, 0.2000],
[5.4000, 3.4000, 1.5000, 0.4000],
[5.3000, 3.7000, 1.5000, 0.2000],
[7.1000, 3.0000, 5.9000, 2.1000],
[4.8000, 3.4000, 1.6000, 0.2000]], dtype=torch.float64), tensor([2, 2, 1, 2, 2, 3, 2, 2, 3, 2, 1, 1, 1, 3, 1])]
2 [tensor([[4.9000, 2.4000, 3.3000, 1.0000],
[5.0000, 2.3000, 3.3000, 1.0000],
[5.7000, 4.4000, 1.5000, 0.4000],
[5.7000, 3.0000, 4.2000, 1.2000],
[5.0000, 3.5000, 1.6000, 0.6000],
[6.7000, 3.0000, 5.0000, 1.7000],
[5.6000, 2.9000, 3.6000, 1.3000],
[6.8000, 3.0000, 5.5000, 2.1000],
[4.4000, 2.9000, 1.4000, 0.2000],
[6.7000, 3.0000, 5.2000, 2.3000],
[5.8000, 2.7000, 4.1000, 1.0000],
[5.5000, 2.5000, 4.0000, 1.3000],
[6.4000, 2.8000, 5.6000, 2.1000],
[5.0000, 3.0000, 1.6000, 0.2000],
[6.4000, 3.2000, 4.5000, 1.5000]], dtype=torch.float64), tensor([2, 2, 1, 2, 1, 2, 2, 3, 1, 3, 2, 2, 3, 1, 2])]
3 [tensor([[4.8000, 3.4000, 1.9000, 0.2000],
[4.6000, 3.4000, 1.4000, 0.3000],
[5.4000, 3.4000, 1.7000, 0.2000],
[7.4000, 2.8000, 6.1000, 1.9000],
[4.9000, 3.1000, 1.5000, 0.1000],
[7.7000, 3.0000, 6.1000, 2.3000],
[6.3000, 3.3000, 6.0000, 2.5000],
[6.1000, 2.6000, 5.6000, 1.4000],
[5.0000, 3.4000, 1.6000, 0.4000],
[6.2000, 3.4000, 5.4000, 2.3000],
[4.9000, 3.1000, 1.5000, 0.1000],
[7.7000, 2.6000, 6.9000, 2.3000],
[4.9000, 3.1000, 1.5000, 0.1000],
[5.0000, 2.0000, 3.5000, 1.0000],
[6.4000, 2.9000, 4.3000, 1.3000]], dtype=torch.float64), tensor([1, 1, 1, 3, 1, 3, 3, 3, 1, 3, 1, 3, 1, 2, 2])]
4 [tensor([[5.1000, 3.8000, 1.6000, 0.2000],
[6.0000, 2.7000, 5.1000, 1.6000],
[7.2000, 3.2000, 6.0000, 1.8000],
[5.1000, 3.8000, 1.5000, 0.3000],
[6.7000, 3.3000, 5.7000, 2.5000],
[6.2000, 2.2000, 4.5000, 1.5000],
[5.8000, 2.7000, 5.1000, 1.9000],
[5.6000, 2.8000, 4.9000, 2.0000],
[6.0000, 2.2000, 5.0000, 1.5000],
[5.1000, 2.5000, 3.0000, 1.1000],
[4.3000, 3.0000, 1.1000, 0.1000],
[6.7000, 3.1000, 4.4000, 1.4000],
[6.5000, 3.2000, 5.1000, 2.0000],
[5.6000, 3.0000, 4.1000, 1.3000],
[5.1000, 3.5000, 1.4000, 0.3000]], dtype=torch.float64), tensor([1, 2, 3, 1, 3, 2, 3, 3, 3, 2, 1, 2, 3, 2, 1])]
5 [tensor([[5.8000, 2.7000, 3.9000, 1.2000],
[7.9000, 3.8000, 6.4000, 2.0000],
[6.0000, 2.2000, 4.0000, 1.0000],
[7.7000, 2.8000, 6.7000, 2.0000],
[5.9000, 3.0000, 5.1000, 1.8000],
[5.8000, 2.8000, 5.1000, 2.4000],
[6.1000, 2.8000, 4.0000, 1.3000],
[5.5000, 4.2000, 1.4000, 0.2000],
[6.2000, 2.9000, 4.3000, 1.3000],
[4.7000, 3.2000, 1.3000, 0.2000],
[7.3000, 2.9000, 6.3000, 1.8000],
[6.3000, 2.3000, 4.4000, 1.3000],
[6.0000, 3.4000, 4.5000, 1.6000],
[6.9000, 3.1000, 5.4000, 2.1000],
[5.0000, 3.4000, 1.5000, 0.2000]], dtype=torch.float64), tensor([2, 3, 2, 3, 3, 3, 2, 1, 2, 1, 3, 2, 2, 3, 1])]
6 [tensor([[4.8000, 3.1000, 1.6000, 0.2000],
[6.3000, 3.3000, 4.7000, 1.6000],
[6.3000, 2.8000, 5.1000, 1.5000],
[4.6000, 3.2000, 1.4000, 0.2000],
[5.9000, 3.0000, 4.2000, 1.5000],
[5.0000, 3.3000, 1.4000, 0.2000],
[5.5000, 3.5000, 1.3000, 0.2000],
[6.4000, 3.1000, 5.5000, 1.8000],
[6.4000, 3.2000, 5.3000, 2.3000],
[5.7000, 2.5000, 5.0000, 2.0000],
[5.4000, 3.9000, 1.3000, 0.4000],
[5.2000, 4.1000, 1.5000, 0.1000],
[6.1000, 3.0000, 4.6000, 1.4000],
[5.0000, 3.5000, 1.3000, 0.3000],
[4.8000, 3.0000, 1.4000, 0.1000]], dtype=torch.float64), tensor([1, 2, 3, 1, 2, 1, 1, 3, 3, 3, 1, 1, 2, 1, 1])]
7 [tensor([[6.1000, 2.8000, 4.7000, 1.2000],
[6.9000, 3.1000, 4.9000, 1.5000],
[5.5000, 2.4000, 3.7000, 1.0000],
[6.2000, 2.8000, 4.8000, 1.8000],
[4.7000, 3.2000, 1.6000, 0.2000],
[4.5000, 2.3000, 1.3000, 0.3000],
[5.0000, 3.2000, 1.2000, 0.2000],
[5.4000, 3.0000, 4.5000, 1.5000],
[5.2000, 3.4000, 1.4000, 0.2000],
[6.7000, 3.3000, 5.7000, 2.1000],
[5.1000, 3.3000, 1.7000, 0.5000],
[4.9000, 3.0000, 1.4000, 0.2000],
[6.5000, 3.0000, 5.5000, 1.8000],
[6.5000, 3.0000, 5.2000, 2.0000],
[6.6000, 2.9000, 4.6000, 1.3000]], dtype=torch.float64), tensor([2, 2, 2, 3, 1, 1, 1, 2, 1, 3, 1, 1, 3, 3, 2])]
8 [tensor([[6.7000, 3.1000, 4.7000, 1.5000],
[7.6000, 3.0000, 6.6000, 2.1000],
[7.2000, 3.6000, 6.1000, 2.5000],
[5.0000, 3.6000, 1.4000, 0.2000],
[6.7000, 3.1000, 5.6000, 2.4000],
[5.2000, 2.7000, 3.9000, 1.4000],
[5.1000, 3.7000, 1.5000, 0.4000],
[6.7000, 2.5000, 5.8000, 1.8000],
[5.9000, 3.2000, 4.8000, 1.8000],
[6.6000, 3.0000, 4.4000, 1.4000],
[7.7000, 3.8000, 6.7000, 2.2000],
[4.6000, 3.6000, 1.0000, 0.2000],
[4.8000, 3.0000, 1.4000, 0.3000],
[4.4000, 3.2000, 1.3000, 0.2000],
[6.3000, 2.7000, 4.9000, 1.8000]], dtype=torch.float64), tensor([2, 3, 3, 1, 3, 2, 1, 3, 2, 2, 3, 1, 1, 1, 3])]
9 [tensor([[5.7000, 3.8000, 1.7000, 0.3000],
[5.7000, 2.8000, 4.5000, 1.3000],
[6.9000, 3.2000, 5.7000, 2.3000],
[5.7000, 2.6000, 3.5000, 1.0000],
[4.9000, 2.5000, 4.5000, 1.7000],
[5.8000, 2.6000, 4.0000, 1.2000],
[6.3000, 2.9000, 5.6000, 1.8000],
[5.8000, 4.0000, 1.2000, 0.2000],
[5.1000, 3.4000, 1.5000, 0.2000],
[5.2000, 3.5000, 1.5000, 0.2000],
[6.3000, 2.5000, 4.9000, 1.5000],
[4.6000, 3.1000, 1.5000, 0.2000],
[6.4000, 2.7000, 5.3000, 1.9000],
[6.8000, 2.8000, 4.8000, 1.4000],
[5.8000, 2.7000, 5.1000, 1.9000]], dtype=torch.float64), tensor([1, 2, 3, 2, 3, 2, 3, 1, 1, 1, 2, 1, 3, 2, 3])]
next(iter(iris_loader))
[tensor([[6.7000, 2.5000, 5.8000, 1.8000],
[6.1000, 2.9000, 4.7000, 1.4000],
[5.3000, 3.7000, 1.5000, 0.2000],
[6.8000, 2.8000, 4.8000, 1.4000],
[6.4000, 3.2000, 5.3000, 2.3000],
[6.0000, 2.7000, 5.1000, 1.6000],
[5.2000, 3.4000, 1.4000, 0.2000],
[5.4000, 3.9000, 1.7000, 0.4000],
[7.4000, 2.8000, 6.1000, 1.9000],
[4.8000, 3.0000, 1.4000, 0.3000],
[5.4000, 3.4000, 1.7000, 0.2000],
[5.5000, 4.2000, 1.4000, 0.2000],
[6.9000, 3.1000, 5.4000, 2.1000],
[5.7000, 2.9000, 4.2000, 1.3000],
[5.1000, 3.5000, 1.4000, 0.2000]], dtype=torch.float64),
tensor([3, 2, 1, 2, 3, 2, 1, 1, 3, 1, 1, 1, 3, 2, 1])]
len(list(iris_loader))
10
print(list(iris_loader))
[[tensor([[4.9000, 3.0000, 1.4000, 0.2000],
[5.7000, 2.8000, 4.5000, 1.3000],
[4.7000, 3.2000, 1.6000, 0.2000],
[6.3000, 2.3000, 4.4000, 1.3000],
[6.4000, 2.8000, 5.6000, 2.2000],
[6.1000, 2.8000, 4.0000, 1.3000],
[6.3000, 2.8000, 5.1000, 1.5000],
[5.8000, 4.0000, 1.2000, 0.2000],
[5.9000, 3.0000, 4.2000, 1.5000],
[5.0000, 3.3000, 1.4000, 0.2000],
[6.5000, 2.8000, 4.6000, 1.5000],
[5.3000, 3.7000, 1.5000, 0.2000],
[6.7000, 3.1000, 5.6000, 2.4000],
[7.2000, 3.6000, 6.1000, 2.5000],
[5.0000, 2.3000, 3.3000, 1.0000]], dtype=torch.float64), tensor([1, 2, 1, 2, 3, 2, 3, 1, 2, 1, 2, 1, 3, 3, 2])], [tensor([[5.7000, 2.9000, 4.2000, 1.3000],
[5.5000, 2.4000, 3.8000, 1.1000],
[6.3000, 3.3000, 6.0000, 2.5000],
[6.0000, 2.2000, 4.0000, 1.0000],
[5.8000, 2.8000, 5.1000, 2.4000],
[6.2000, 3.4000, 5.4000, 2.3000],
[4.6000, 3.4000, 1.4000, 0.3000],
[6.8000, 3.2000, 5.9000, 2.3000],
[5.0000, 3.6000, 1.4000, 0.2000],
[4.9000, 3.1000, 1.5000, 0.1000],
[5.6000, 2.9000, 3.6000, 1.3000],
[6.7000, 3.1000, 4.4000, 1.4000],
[6.1000, 3.0000, 4.6000, 1.4000],
[6.3000, 2.9000, 5.6000, 1.8000],
[7.3000, 2.9000, 6.3000, 1.8000]], dtype=torch.float64), tensor([2, 2, 3, 2, 3, 3, 1, 3, 1, 1, 2, 2, 2, 3, 3])], [tensor([[6.1000, 2.8000, 4.7000, 1.2000],
[6.7000, 3.1000, 4.7000, 1.5000],
[4.8000, 3.0000, 1.4000, 0.1000],
[4.9000, 3.1000, 1.5000, 0.1000],
[6.3000, 2.5000, 5.0000, 1.9000],
[5.7000, 4.4000, 1.5000, 0.4000],
[7.9000, 3.8000, 6.4000, 2.0000],
[5.2000, 3.5000, 1.5000, 0.2000],
[6.5000, 3.0000, 5.2000, 2.0000],
[5.6000, 3.0000, 4.1000, 1.3000],
[5.1000, 3.4000, 1.5000, 0.2000],
[5.8000, 2.7000, 4.1000, 1.0000],
[4.9000, 3.1000, 1.5000, 0.1000],
[6.0000, 2.2000, 5.0000, 1.5000],
[6.4000, 3.1000, 5.5000, 1.8000]], dtype=torch.float64), tensor([2, 2, 1, 1, 3, 1, 3, 1, 3, 2, 1, 2, 1, 3, 3])], [tensor([[4.3000, 3.0000, 1.1000, 0.1000],
[6.3000, 3.4000, 5.6000, 2.4000],
[7.4000, 2.8000, 6.1000, 1.9000],
[5.4000, 3.7000, 1.5000, 0.2000],
[6.4000, 2.7000, 5.3000, 1.9000],
[5.0000, 3.5000, 1.6000, 0.6000],
[5.6000, 2.7000, 4.2000, 1.3000],
[7.1000, 3.0000, 5.9000, 2.1000],
[4.5000, 2.3000, 1.3000, 0.3000],
[5.0000, 2.0000, 3.5000, 1.0000],
[5.6000, 3.0000, 4.5000, 1.5000],
[4.4000, 2.9000, 1.4000, 0.2000],
[6.6000, 2.9000, 4.6000, 1.3000],
[6.4000, 3.2000, 5.3000, 2.3000],
[5.5000, 4.2000, 1.4000, 0.2000]], dtype=torch.float64), tensor([1, 3, 3, 1, 3, 1, 2, 3, 1, 2, 2, 1, 2, 3, 1])], [tensor([[5.0000, 3.0000, 1.6000, 0.2000],
[5.4000, 3.9000, 1.7000, 0.4000],
[4.9000, 2.5000, 4.5000, 1.7000],
[4.9000, 2.4000, 3.3000, 1.0000],
[5.1000, 3.5000, 1.4000, 0.2000],
[5.1000, 3.8000, 1.5000, 0.3000],
[4.6000, 3.1000, 1.5000, 0.2000],
[5.4000, 3.4000, 1.5000, 0.4000],
[5.6000, 2.5000, 3.9000, 1.1000],
[6.8000, 2.8000, 4.8000, 1.4000],
[4.4000, 3.0000, 1.3000, 0.2000],
[6.0000, 3.4000, 4.5000, 1.6000],
[6.7000, 3.0000, 5.2000, 2.3000],
[5.8000, 2.7000, 5.1000, 1.9000],
[6.0000, 2.9000, 4.5000, 1.5000]], dtype=torch.float64), tensor([1, 1, 3, 2, 1, 1, 1, 1, 2, 2, 1, 2, 3, 3, 2])], [tensor([[5.0000, 3.4000, 1.5000, 0.2000],
[4.7000, 3.2000, 1.3000, 0.2000],
[6.9000, 3.1000, 5.4000, 2.1000],
[6.7000, 2.5000, 5.8000, 1.8000],
[5.7000, 2.8000, 4.1000, 1.3000],
[4.8000, 3.4000, 1.9000, 0.2000],
[5.5000, 2.3000, 4.0000, 1.3000],
[7.7000, 2.6000, 6.9000, 2.3000],
[6.3000, 3.3000, 4.7000, 1.6000],
[6.6000, 3.0000, 4.4000, 1.4000],
[5.7000, 3.8000, 1.7000, 0.3000],
[5.8000, 2.7000, 3.9000, 1.2000],
[5.1000, 2.5000, 3.0000, 1.1000],
[4.4000, 3.2000, 1.3000, 0.2000],
[5.0000, 3.5000, 1.3000, 0.3000]], dtype=torch.float64), tensor([1, 1, 3, 3, 2, 1, 2, 3, 2, 2, 1, 2, 2, 1, 1])], [tensor([[5.1000, 3.3000, 1.7000, 0.5000],
[5.1000, 3.5000, 1.4000, 0.3000],
[6.1000, 3.0000, 4.9000, 1.8000],
[6.0000, 2.7000, 5.1000, 1.6000],
[5.5000, 2.4000, 3.7000, 1.0000],
[6.9000, 3.1000, 5.1000, 2.3000],
[6.4000, 3.2000, 4.5000, 1.5000],
[5.8000, 2.7000, 5.1000, 1.9000],
[6.5000, 3.0000, 5.5000, 1.8000],
[5.4000, 3.4000, 1.7000, 0.2000],
[6.5000, 3.2000, 5.1000, 2.0000],
[6.2000, 2.2000, 4.5000, 1.5000],
[7.2000, 3.0000, 5.8000, 1.6000],
[5.2000, 4.1000, 1.5000, 0.1000],
[6.0000, 3.0000, 4.8000, 1.8000]], dtype=torch.float64), tensor([1, 1, 3, 2, 2, 3, 2, 3, 3, 1, 3, 2, 3, 1, 3])], [tensor([[5.7000, 2.5000, 5.0000, 2.0000],
[5.1000, 3.7000, 1.5000, 0.4000],
[5.4000, 3.9000, 1.3000, 0.4000],
[6.2000, 2.8000, 4.8000, 1.8000],
[5.7000, 2.6000, 3.5000, 1.0000],
[7.7000, 2.8000, 6.7000, 2.0000],
[5.8000, 2.6000, 4.0000, 1.2000],
[7.2000, 3.2000, 6.0000, 1.8000],
[4.8000, 3.1000, 1.6000, 0.2000],
[6.2000, 2.9000, 4.3000, 1.3000],
[7.7000, 3.0000, 6.1000, 2.3000],
[5.4000, 3.0000, 4.5000, 1.5000],
[6.7000, 3.3000, 5.7000, 2.5000],
[5.1000, 3.8000, 1.9000, 0.4000],
[6.1000, 2.6000, 5.6000, 1.4000]], dtype=torch.float64), tensor([3, 1, 1, 3, 2, 3, 2, 3, 1, 2, 3, 2, 3, 1, 3])], [tensor([[6.4000, 2.8000, 5.6000, 2.1000],
[5.5000, 3.5000, 1.3000, 0.2000],
[4.8000, 3.4000, 1.6000, 0.2000],
[5.9000, 3.0000, 5.1000, 1.8000],
[5.1000, 3.8000, 1.6000, 0.2000],
[6.4000, 2.9000, 4.3000, 1.3000],
[5.5000, 2.5000, 4.0000, 1.3000],
[5.2000, 3.4000, 1.4000, 0.2000],
[6.9000, 3.2000, 5.7000, 2.3000],
[6.7000, 3.0000, 5.0000, 1.7000],
[6.3000, 2.7000, 4.9000, 1.8000],
[6.8000, 3.0000, 5.5000, 2.1000],
[5.9000, 3.2000, 4.8000, 1.8000],
[4.6000, 3.2000, 1.4000, 0.2000],
[5.6000, 2.8000, 4.9000, 2.0000]], dtype=torch.float64), tensor([3, 1, 1, 3, 1, 2, 2, 1, 3, 2, 3, 3, 2, 1, 3])], [tensor([[5.0000, 3.2000, 1.2000, 0.2000],
[7.6000, 3.0000, 6.6000, 2.1000],
[6.7000, 3.3000, 5.7000, 2.1000],
[6.5000, 3.0000, 5.8000, 2.2000],
[4.8000, 3.0000, 1.4000, 0.3000],
[6.1000, 2.9000, 4.7000, 1.4000],
[7.7000, 3.8000, 6.7000, 2.2000],
[4.6000, 3.6000, 1.0000, 0.2000],
[5.5000, 2.6000, 4.4000, 1.2000],
[7.0000, 3.2000, 4.7000, 1.4000],
[5.0000, 3.4000, 1.6000, 0.4000],
[5.7000, 3.0000, 4.2000, 1.2000],
[6.9000, 3.1000, 4.9000, 1.5000],
[5.2000, 2.7000, 3.9000, 1.4000],
[6.3000, 2.5000, 4.9000, 1.5000]], dtype=torch.float64), tensor([1, 3, 3, 3, 1, 2, 3, 1, 2, 2, 1, 2, 2, 2, 2])]]
len((list(iris_loader)))
10
print((iris_loader))
<torch.utils.data.dataloader.DataLoader object at 0x7fad2ac187f0>
A Quick Note on Torchvision
PyTorch offers another powerful dataset tool called torchvision, which is useful when working with image data. We’ll go into a lot more detail in the Convolutional Neural Network (CNN) section. For now, just know that torchvision offers built-in image datasets like MNIST and CIFAR-10, as well as tools for transforming images into tensors.
Creating a Custom Dataset for your files
A custom Dataset class must implement three functions: init, len, and getitem.
lets build custom dataset for iris
data_path = '/content/IRIS.csv'
from torch.utils.data import Dataset
class CustomImageDataset(Dataset):
def __init__(self, data_path):
self.data = pd.read_csv(data_path)
def __len__(self):
return len(self.data)
def __getitem__(self, idx):
record = self.data.iloc[idx]
features = list(record.drop(index=['species']).values)
features = torch.tensor(features)
label_name = record['species']
repl = {'Iris-setosa':1, 'Iris-versicolor':2 , 'Iris-virginica':3}
label = repl[label_name]
label = torch.tensor([label])
return features, label
iris = CustomImageDataset(data_path)
len(iris)
150
type(iris)
__main__.CustomImageDataset
for i in iris:
print(i)
(tensor([5.1000, 3.5000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.9000, 3.0000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.7000, 3.2000, 1.3000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.6000, 3.1000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.6000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.4000, 3.9000, 1.7000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([4.6000, 3.4000, 1.4000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.4000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.4000, 2.9000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([5.4000, 3.7000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.8000, 3.4000, 1.6000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.8000, 3.0000, 1.4000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([4.3000, 3.0000, 1.1000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([5.8000, 4.0000, 1.2000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.7000, 4.4000, 1.5000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([5.4000, 3.9000, 1.3000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.5000, 1.4000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([5.7000, 3.8000, 1.7000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.8000, 1.5000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([5.4000, 3.4000, 1.7000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.7000, 1.5000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([4.6000, 3.6000, 1.0000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.3000, 1.7000, 0.5000], dtype=torch.float64), tensor([1]))
(tensor([4.8000, 3.4000, 1.9000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.0000, 1.6000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.4000, 1.6000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([5.2000, 3.5000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.2000, 3.4000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.7000, 3.2000, 1.6000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.8000, 3.1000, 1.6000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.4000, 3.4000, 1.5000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([5.2000, 4.1000, 1.5000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([5.5000, 4.2000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.2000, 1.2000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.5000, 3.5000, 1.3000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.9000, 3.1000, 1.5000, 0.1000], dtype=torch.float64), tensor([1]))
(tensor([4.4000, 3.0000, 1.3000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.4000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.5000, 1.3000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([4.5000, 2.3000, 1.3000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([4.4000, 3.2000, 1.3000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.5000, 1.6000, 0.6000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.8000, 1.9000, 0.4000], dtype=torch.float64), tensor([1]))
(tensor([4.8000, 3.0000, 1.4000, 0.3000], dtype=torch.float64), tensor([1]))
(tensor([5.1000, 3.8000, 1.6000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([4.6000, 3.2000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.3000, 3.7000, 1.5000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([5.0000, 3.3000, 1.4000, 0.2000], dtype=torch.float64), tensor([1]))
(tensor([7.0000, 3.2000, 4.7000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([6.4000, 3.2000, 4.5000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([6.9000, 3.1000, 4.9000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([5.5000, 2.3000, 4.0000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.5000, 2.8000, 4.6000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([5.7000, 2.8000, 4.5000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.3000, 3.3000, 4.7000, 1.6000], dtype=torch.float64), tensor([2]))
(tensor([4.9000, 2.4000, 3.3000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([6.6000, 2.9000, 4.6000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.2000, 2.7000, 3.9000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([5.0000, 2.0000, 3.5000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([5.9000, 3.0000, 4.2000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([6.0000, 2.2000, 4.0000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([6.1000, 2.9000, 4.7000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([5.6000, 2.9000, 3.6000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.7000, 3.1000, 4.4000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([5.6000, 3.0000, 4.5000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([5.8000, 2.7000, 4.1000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([6.2000, 2.2000, 4.5000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([5.6000, 2.5000, 3.9000, 1.1000], dtype=torch.float64), tensor([2]))
(tensor([5.9000, 3.2000, 4.8000, 1.8000], dtype=torch.float64), tensor([2]))
(tensor([6.1000, 2.8000, 4.0000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.3000, 2.5000, 4.9000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([6.1000, 2.8000, 4.7000, 1.2000], dtype=torch.float64), tensor([2]))
(tensor([6.4000, 2.9000, 4.3000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.6000, 3.0000, 4.4000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([6.8000, 2.8000, 4.8000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([6.7000, 3.0000, 5.0000, 1.7000], dtype=torch.float64), tensor([2]))
(tensor([6.0000, 2.9000, 4.5000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([5.7000, 2.6000, 3.5000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([5.5000, 2.4000, 3.8000, 1.1000], dtype=torch.float64), tensor([2]))
(tensor([5.5000, 2.4000, 3.7000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([5.8000, 2.7000, 3.9000, 1.2000], dtype=torch.float64), tensor([2]))
(tensor([6.0000, 2.7000, 5.1000, 1.6000], dtype=torch.float64), tensor([2]))
(tensor([5.4000, 3.0000, 4.5000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([6.0000, 3.4000, 4.5000, 1.6000], dtype=torch.float64), tensor([2]))
(tensor([6.7000, 3.1000, 4.7000, 1.5000], dtype=torch.float64), tensor([2]))
(tensor([6.3000, 2.3000, 4.4000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.6000, 3.0000, 4.1000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.5000, 2.5000, 4.0000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.5000, 2.6000, 4.4000, 1.2000], dtype=torch.float64), tensor([2]))
(tensor([6.1000, 3.0000, 4.6000, 1.4000], dtype=torch.float64), tensor([2]))
(tensor([5.8000, 2.6000, 4.0000, 1.2000], dtype=torch.float64), tensor([2]))
(tensor([5.0000, 2.3000, 3.3000, 1.0000], dtype=torch.float64), tensor([2]))
(tensor([5.6000, 2.7000, 4.2000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.7000, 3.0000, 4.2000, 1.2000], dtype=torch.float64), tensor([2]))
(tensor([5.7000, 2.9000, 4.2000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.2000, 2.9000, 4.3000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([5.1000, 2.5000, 3.0000, 1.1000], dtype=torch.float64), tensor([2]))
(tensor([5.7000, 2.8000, 4.1000, 1.3000], dtype=torch.float64), tensor([2]))
(tensor([6.3000, 3.3000, 6.0000, 2.5000], dtype=torch.float64), tensor([3]))
(tensor([5.8000, 2.7000, 5.1000, 1.9000], dtype=torch.float64), tensor([3]))
(tensor([7.1000, 3.0000, 5.9000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([6.3000, 2.9000, 5.6000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.5000, 3.0000, 5.8000, 2.2000], dtype=torch.float64), tensor([3]))
(tensor([7.6000, 3.0000, 6.6000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([4.9000, 2.5000, 4.5000, 1.7000], dtype=torch.float64), tensor([3]))
(tensor([7.3000, 2.9000, 6.3000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.7000, 2.5000, 5.8000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([7.2000, 3.6000, 6.1000, 2.5000], dtype=torch.float64), tensor([3]))
(tensor([6.5000, 3.2000, 5.1000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([6.4000, 2.7000, 5.3000, 1.9000], dtype=torch.float64), tensor([3]))
(tensor([6.8000, 3.0000, 5.5000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([5.7000, 2.5000, 5.0000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([5.8000, 2.8000, 5.1000, 2.4000], dtype=torch.float64), tensor([3]))
(tensor([6.4000, 3.2000, 5.3000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([6.5000, 3.0000, 5.5000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([7.7000, 3.8000, 6.7000, 2.2000], dtype=torch.float64), tensor([3]))
(tensor([7.7000, 2.6000, 6.9000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([6.0000, 2.2000, 5.0000, 1.5000], dtype=torch.float64), tensor([3]))
(tensor([6.9000, 3.2000, 5.7000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([5.6000, 2.8000, 4.9000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([7.7000, 2.8000, 6.7000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([6.3000, 2.7000, 4.9000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.7000, 3.3000, 5.7000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([7.2000, 3.2000, 6.0000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.2000, 2.8000, 4.8000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.1000, 3.0000, 4.9000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.4000, 2.8000, 5.6000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([7.2000, 3.0000, 5.8000, 1.6000], dtype=torch.float64), tensor([3]))
(tensor([7.4000, 2.8000, 6.1000, 1.9000], dtype=torch.float64), tensor([3]))
(tensor([7.9000, 3.8000, 6.4000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([6.4000, 2.8000, 5.6000, 2.2000], dtype=torch.float64), tensor([3]))
(tensor([6.3000, 2.8000, 5.1000, 1.5000], dtype=torch.float64), tensor([3]))
(tensor([6.1000, 2.6000, 5.6000, 1.4000], dtype=torch.float64), tensor([3]))
(tensor([7.7000, 3.0000, 6.1000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([6.3000, 3.4000, 5.6000, 2.4000], dtype=torch.float64), tensor([3]))
(tensor([6.4000, 3.1000, 5.5000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.0000, 3.0000, 4.8000, 1.8000], dtype=torch.float64), tensor([3]))
(tensor([6.9000, 3.1000, 5.4000, 2.1000], dtype=torch.float64), tensor([3]))
(tensor([6.7000, 3.1000, 5.6000, 2.4000], dtype=torch.float64), tensor([3]))
(tensor([6.9000, 3.1000, 5.1000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([5.8000, 2.7000, 5.1000, 1.9000], dtype=torch.float64), tensor([3]))
(tensor([6.8000, 3.2000, 5.9000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([6.7000, 3.3000, 5.7000, 2.5000], dtype=torch.float64), tensor([3]))
(tensor([6.7000, 3.0000, 5.2000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([6.3000, 2.5000, 5.0000, 1.9000], dtype=torch.float64), tensor([3]))
(tensor([6.5000, 3.0000, 5.2000, 2.0000], dtype=torch.float64), tensor([3]))
(tensor([6.2000, 3.4000, 5.4000, 2.3000], dtype=torch.float64), tensor([3]))
(tensor([5.9000, 3.0000, 5.1000, 1.8000], dtype=torch.float64), tensor([3]))
iris_loader = DataLoader(iris, batch_size=15, shuffle=True)
print(iris_loader)
for i_batch, sample_batched in enumerate(iris_loader):
print(i_batch, sample_batched)
0 [tensor([[7.0000, 3.2000, 4.7000, 1.4000],
[5.6000, 3.0000, 4.5000, 1.5000],
[6.8000, 3.0000, 5.5000, 2.1000],
[6.1000, 2.8000, 4.0000, 1.3000],
[6.2000, 2.9000, 4.3000, 1.3000],
[5.8000, 2.6000, 4.0000, 1.2000],
[4.9000, 3.1000, 1.5000, 0.1000],
[5.6000, 2.8000, 4.9000, 2.0000],
[6.6000, 3.0000, 4.4000, 1.4000],
[7.4000, 2.8000, 6.1000, 1.9000],
[6.7000, 3.0000, 5.2000, 2.3000],
[5.0000, 3.5000, 1.6000, 0.6000],
[5.7000, 2.8000, 4.5000, 1.3000],
[5.5000, 2.5000, 4.0000, 1.3000],
[6.3000, 2.5000, 4.9000, 1.5000]], dtype=torch.float64), tensor([2, 2, 3, 2, 2, 2, 1, 3, 2, 3, 3, 1, 2, 2, 2])]
1 [tensor([[5.7000, 2.9000, 4.2000, 1.3000],
[6.9000, 3.1000, 5.1000, 2.3000],
[4.9000, 3.0000, 1.4000, 0.2000],
[6.3000, 3.3000, 6.0000, 2.5000],
[6.7000, 2.5000, 5.8000, 1.8000],
[5.5000, 2.4000, 3.7000, 1.0000],
[6.3000, 2.7000, 4.9000, 1.8000],
[6.0000, 2.9000, 4.5000, 1.5000],
[5.1000, 3.8000, 1.9000, 0.4000],
[4.8000, 3.0000, 1.4000, 0.1000],
[6.3000, 2.3000, 4.4000, 1.3000],
[7.7000, 2.6000, 6.9000, 2.3000],
[5.5000, 4.2000, 1.4000, 0.2000],
[5.6000, 2.7000, 4.2000, 1.3000],
[6.3000, 2.8000, 5.1000, 1.5000]], dtype=torch.float64), tensor([2, 3, 1, 3, 3, 2, 3, 2, 1, 1, 2, 3, 1, 2, 3])]
2 [tensor([[6.3000, 3.3000, 4.7000, 1.6000],
[5.6000, 3.0000, 4.1000, 1.3000],
[5.7000, 4.4000, 1.5000, 0.4000],
[6.2000, 3.4000, 5.4000, 2.3000],
[6.8000, 2.8000, 4.8000, 1.4000],
[5.5000, 2.6000, 4.4000, 1.2000],
[6.4000, 2.8000, 5.6000, 2.2000],
[5.1000, 3.3000, 1.7000, 0.5000],
[5.0000, 3.3000, 1.4000, 0.2000],
[4.8000, 3.1000, 1.6000, 0.2000],
[4.6000, 3.6000, 1.0000, 0.2000],
[6.3000, 2.5000, 5.0000, 1.9000],
[6.0000, 2.7000, 5.1000, 1.6000],
[6.7000, 3.1000, 5.6000, 2.4000],
[6.5000, 2.8000, 4.6000, 1.5000]], dtype=torch.float64), tensor([2, 2, 1, 3, 2, 2, 3, 1, 1, 1, 1, 3, 2, 3, 2])]
3 [tensor([[4.6000, 3.1000, 1.5000, 0.2000],
[6.8000, 3.2000, 5.9000, 2.3000],
[5.9000, 3.0000, 5.1000, 1.8000],
[6.4000, 3.2000, 4.5000, 1.5000],
[4.6000, 3.4000, 1.4000, 0.3000],
[4.9000, 3.1000, 1.5000, 0.1000],
[4.9000, 2.4000, 3.3000, 1.0000],
[5.0000, 3.0000, 1.6000, 0.2000],
[5.4000, 3.7000, 1.5000, 0.2000],
[7.7000, 3.0000, 6.1000, 2.3000],
[7.7000, 2.8000, 6.7000, 2.0000],
[6.9000, 3.1000, 4.9000, 1.5000],
[5.6000, 2.5000, 3.9000, 1.1000],
[6.4000, 3.1000, 5.5000, 1.8000],
[6.0000, 3.0000, 4.8000, 1.8000]], dtype=torch.float64), tensor([1, 3, 3, 2, 1, 1, 2, 1, 1, 3, 3, 2, 2, 3, 3])]
4 [tensor([[5.9000, 3.0000, 4.2000, 1.5000],
[4.4000, 3.2000, 1.3000, 0.2000],
[5.7000, 2.6000, 3.5000, 1.0000],
[5.1000, 3.8000, 1.6000, 0.2000],
[6.7000, 3.3000, 5.7000, 2.1000],
[6.9000, 3.1000, 5.4000, 2.1000],
[5.5000, 2.4000, 3.8000, 1.1000],
[7.1000, 3.0000, 5.9000, 2.1000],
[6.1000, 2.6000, 5.6000, 1.4000],
[5.1000, 3.5000, 1.4000, 0.3000],
[5.4000, 3.9000, 1.3000, 0.4000],
[6.4000, 2.8000, 5.6000, 2.1000],
[6.1000, 3.0000, 4.9000, 1.8000],
[5.4000, 3.4000, 1.5000, 0.4000],
[6.7000, 3.0000, 5.0000, 1.7000]], dtype=torch.float64), tensor([2, 1, 2, 1, 3, 3, 2, 3, 3, 1, 1, 3, 3, 1, 2])]
5 [tensor([[5.0000, 3.4000, 1.6000, 0.4000],
[6.0000, 2.2000, 5.0000, 1.5000],
[7.3000, 2.9000, 6.3000, 1.8000],
[5.0000, 2.3000, 3.3000, 1.0000],
[5.7000, 3.8000, 1.7000, 0.3000],
[7.2000, 3.2000, 6.0000, 1.8000],
[5.5000, 3.5000, 1.3000, 0.2000],
[4.8000, 3.4000, 1.9000, 0.2000],
[5.1000, 3.5000, 1.4000, 0.2000],
[5.0000, 3.5000, 1.3000, 0.3000],
[4.4000, 3.0000, 1.3000, 0.2000],
[5.1000, 3.7000, 1.5000, 0.4000],
[5.8000, 2.7000, 4.1000, 1.0000],
[4.5000, 2.3000, 1.3000, 0.3000],
[5.2000, 3.5000, 1.5000, 0.2000]], dtype=torch.float64), tensor([1, 3, 3, 2, 1, 3, 1, 1, 1, 1, 1, 1, 2, 1, 1])]
6 [tensor([[4.7000, 3.2000, 1.3000, 0.2000],
[6.9000, 3.2000, 5.7000, 2.3000],
[6.7000, 3.3000, 5.7000, 2.5000],
[5.1000, 3.4000, 1.5000, 0.2000],
[4.9000, 3.1000, 1.5000, 0.1000],
[7.2000, 3.6000, 6.1000, 2.5000],
[5.0000, 3.6000, 1.4000, 0.2000],
[6.5000, 3.2000, 5.1000, 2.0000],
[6.5000, 3.0000, 5.8000, 2.2000],
[4.8000, 3.4000, 1.6000, 0.2000],
[4.8000, 3.0000, 1.4000, 0.3000],
[6.4000, 3.2000, 5.3000, 2.3000],
[4.3000, 3.0000, 1.1000, 0.1000],
[6.4000, 2.7000, 5.3000, 1.9000],
[6.5000, 3.0000, 5.5000, 1.8000]], dtype=torch.float64), tensor([1, 3, 3, 1, 1, 3, 1, 3, 3, 1, 1, 3, 1, 3, 3])]
7 [tensor([[5.2000, 2.7000, 3.9000, 1.4000],
[5.8000, 2.8000, 5.1000, 2.4000],
[5.4000, 3.4000, 1.7000, 0.2000],
[5.0000, 3.2000, 1.2000, 0.2000],
[5.6000, 2.9000, 3.6000, 1.3000],
[5.2000, 3.4000, 1.4000, 0.2000],
[6.7000, 3.1000, 4.4000, 1.4000],
[5.1000, 2.5000, 3.0000, 1.1000],
[6.4000, 2.9000, 4.3000, 1.3000],
[5.8000, 4.0000, 1.2000, 0.2000],
[5.3000, 3.7000, 1.5000, 0.2000],
[7.2000, 3.0000, 5.8000, 1.6000],
[6.0000, 2.2000, 4.0000, 1.0000],
[4.9000, 2.5000, 4.5000, 1.7000],
[4.4000, 2.9000, 1.4000, 0.2000]], dtype=torch.float64), tensor([2, 3, 1, 1, 2, 1, 2, 2, 2, 1, 1, 3, 2, 3, 1])]
8 [tensor([[6.7000, 3.1000, 4.7000, 1.5000],
[5.4000, 3.0000, 4.5000, 1.5000],
[5.7000, 2.5000, 5.0000, 2.0000],
[5.8000, 2.7000, 5.1000, 1.9000],
[7.7000, 3.8000, 6.7000, 2.2000],
[5.8000, 2.7000, 5.1000, 1.9000],
[6.1000, 2.9000, 4.7000, 1.4000],
[5.8000, 2.7000, 3.9000, 1.2000],
[6.0000, 3.4000, 4.5000, 1.6000],
[6.1000, 3.0000, 4.6000, 1.4000],
[4.7000, 3.2000, 1.6000, 0.2000],
[6.2000, 2.8000, 4.8000, 1.8000],
[6.3000, 2.9000, 5.6000, 1.8000],
[6.6000, 2.9000, 4.6000, 1.3000],
[6.1000, 2.8000, 4.7000, 1.2000]], dtype=torch.float64), tensor([2, 2, 3, 3, 3, 3, 2, 2, 2, 2, 1, 3, 3, 2, 2])]
9 [tensor([[5.7000, 3.0000, 4.2000, 1.2000],
[7.9000, 3.8000, 6.4000, 2.0000],
[7.6000, 3.0000, 6.6000, 2.1000],
[6.5000, 3.0000, 5.2000, 2.0000],
[5.0000, 3.4000, 1.5000, 0.2000],
[5.4000, 3.9000, 1.7000, 0.4000],
[6.2000, 2.2000, 4.5000, 1.5000],
[5.0000, 2.0000, 3.5000, 1.0000],
[5.7000, 2.8000, 4.1000, 1.3000],
[5.5000, 2.3000, 4.0000, 1.3000],
[5.1000, 3.8000, 1.5000, 0.3000],
[5.2000, 4.1000, 1.5000, 0.1000],
[6.3000, 3.4000, 5.6000, 2.4000],
[4.6000, 3.2000, 1.4000, 0.2000],
[5.9000, 3.2000, 4.8000, 1.8000]], dtype=torch.float64), tensor([2, 3, 3, 3, 1, 1, 2, 2, 2, 2, 1, 1, 3, 1, 2])]