如何创建一个数据集从磁盘上的三个文件与数据集库在python?

Question

磁盘上有三个名为train.xlsx、validation.xlsx和test.xlsx的文件。我需要一个包含这三个文件的数据集库。下面是我的代码:

from google.colab import drive
from datasets import Dataset
import pandas as pd
drive.mount('/content/drive')
train_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/train.xlsx')
validation_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/valid.xlsx')
test_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/test.xlsx')
print(train_data.shape)
print(validation_data.shape)
print(test_data.shape)

现在我需要有一个数据集与这些键从相应的文件:数据集['train']和数据集['validation']和数据集['test']有人能帮我吗?

Answer 1

试试这个

train_data = train_data.values.tolist()
validation_data = validation_data.values.tolist()
test_data = test_data.values.tolist()
d = {'train_data ' : train_data ,
'validation_data ' : validation_data ,
'test_data ' : test_data 
}
df = pd.DataFrame(data = d)

值得注意的是，.values.tolist()工作，如果这些数据帧有一个列，如果没有，指定它EX:train_data ['COLUMN'].values.tolist()

Answer 2

试试

import pandas as pd
import os
from google.colab import drive
drive.mount('/content/drive')
os.chdir('/content/drive/My Drive/NLP-Datasets/Question2_Data/')
train_data = 'train.xlsx'
validation_data = 'valid.xlsx'
test_data = 'test.xlsx'
paths = [train_data, validation_data , test_data ]
dfs = {p: pd.read_excel(p) for p in paths}

更新:您可以使用Python中的datassets库从磁盘上的三个文件创建一个数据集，如下所示:

from google.colab import drive
from datasets import Dataset
import pandas as pd
# Mount Google Drive
drive.mount('/content/drive')
# Load train, validation, and test data
train_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/train.xlsx')
validation_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/valid.xlsx')
test_data = pd.read_excel('/content/drive/My Drive/NLP-Datasets/Question2_Data/test.xlsx')
# Convert data to dictionary format
train_dict = train_data.to_dict(orient='list')
validation_dict = validation_data.to_dict(orient='list')
test_dict = test_data.to_dict(orient='list')
# Create a dataset from the data
dataset = Dataset.from_dict({
'train': train_dict,
'validation': validation_dict,
'test': test_dict
})
# Print the shapes of the data
print(dataset['train'].shape)
print(dataset['validation'].shape)
print(dataset['test'].shape)