ium_z360792/dataset.py

103 lines
2.1 KiB
Python

import os
import zipfile
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
os.system("kaggle datasets download -d gpreda/covid-world-vaccination-progress")
with zipfile.ZipFile("covid-world-vaccination-progress.zip", "r") as zip_ref:
zip_ref.extractall("country_vaccinations")
csv_file = None
for root, dirs, files in os.walk("country_vaccinations"):
for file in files:
if file.endswith(".csv"):
csv_file = os.path.join(root, file)
break
if csv_file is None:
raise FileNotFoundError("CSV file not found in the extracted dataset")
data = pd.read_csv(csv_file)
train_data, temp_data = train_test_split(data, test_size=0.4, random_state=42)
dev_data, test_data = train_test_split(temp_data, test_size=0.5, random_state=42)
def print_stats(df, name):
print(f"\nStatystyki dla {name}:")
print(f"Wielkość zbioru: {len(df)}")
for col in df.columns:
if df[col].dtype != "object":
print(f"\nParametr: {col}")
print(f"Minimum: {df[col].min()}")
print(f"Maksimum: {df[col].max()}")
print(f"Średnia: {df[col].mean()}")
print(f"Odchylenie standardowe: {df[col].std()}")
print(f"Mediana: {df[col].median()}")
print_stats(data, "Cały zbiór")
print_stats(train_data, "Zbiór treningowy")
print_stats(dev_data, "Zbiór walidacyjny")
print_stats(test_data, "Zbiór testowy")
def normalize_data(df, columns):
scaler = MinMaxScaler()
for col in columns:
if df[col].dtype != "object":
df[col] = scaler.fit_transform(df[[col]])
normalize_data(train_data, train_data.columns)
normalize_data(dev_data, dev_data.columns)
normalize_data(test_data, test_data.columns)
def clean_data(df):
df.dropna(inplace=True)
df.drop_duplicates(inplace=True)
clean_data(data)
clean_data(train_data)
clean_data(dev_data)
clean_data(test_data)
train_data.to_csv("./results/train_data.csv", index=False)
dev_data.to_csv("./results/dev_data.csv", index=False)
test_data.to_csv("./results/test_data.csv", index=False)