mirror of
https://github.com/aladdinpersson/Machine-Learning-Collection.git
synced 2026-02-20 13:50:41 +00:00
24 lines
890 B
Python
24 lines
890 B
Python
import pandas as pd
|
|
import torch
|
|
from torch.utils.data import TensorDataset
|
|
from torch.utils.data.dataset import random_split
|
|
from math import ceil
|
|
|
|
def get_data():
|
|
train_data = pd.read_csv("new_shiny_train.csv")
|
|
y = train_data["target"]
|
|
X = train_data.drop(["ID_code", "target"], axis=1)
|
|
X_tensor = torch.tensor(X.values, dtype=torch.float32)
|
|
y_tensor = torch.tensor(y.values, dtype=torch.float32)
|
|
ds = TensorDataset(X_tensor, y_tensor)
|
|
train_ds, val_ds = random_split(ds, [int(0.999*len(ds)), ceil(0.001*len(ds))])
|
|
|
|
test_data = pd.read_csv("new_shiny_test.csv")
|
|
test_ids = test_data["ID_code"]
|
|
X = test_data.drop(["ID_code"], axis=1)
|
|
X_tensor = torch.tensor(X.values, dtype=torch.float32)
|
|
y_tensor = torch.tensor(y.values, dtype=torch.float32)
|
|
test_ds = TensorDataset(X_tensor, y_tensor)
|
|
|
|
return train_ds, val_ds, test_ds, test_ids
|