import pandas as pd from sklearn.model_selection import train_test_split from sklearn.preprocessing import LabelEncoder

Load the dataset

df = pd.read_csv('accidents.csv')

Handle missing values

df.dropna(inplace=True)

Encode categorical data

le = LabelEncoder() df['weather'] = le.fit_transform(df['weather']) df['location'] = le.fit_transform(df['location'])

Define features and target

X = df[['time', 'location', 'weather']] y = df['severity'] # Assume 'severity' is the target to predict

Split into training and testing sets

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)