Transaction data into a one-hot encoded boolean array without using TransactionEncoder

Question

I want to transform a transaction data into a one-hot encoded boolean array without using TransactionEncoder. or to create a function from scratch that do same works as TransactionEncoder. after following code :

    df = pd.read_csv('basket.csv')
    df.set_index('ID', inplace=True)
    transactions = []
    for i in range(0, len(df)): 
        transactions.append([str(df.values[i,j]) for j in range(0, len(df.columns))])
        
    print(transactions[:5])
    
    flattened = [item for transaction in transactions for item in transaction]
    print(len(flattened))

output :

[['whole milk', 'eggs', 'salty snack', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan'], ['whole milk', 'eggs', 'white bread', 'yogurt', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan'], ['whole milk', 'eggs', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan'], ['whole milk', 'eggs', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan'], ['whole milk', 'eggs', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan', 'nan']]