pythonnumpyloopsrandomquantile

How can I create 1000 columns of random variables using each row's mean and SD

My dataframe looks like this:

City Mean SD
Newcastle 60 0.81
Liverpool 62 0.91
Cardiff 65 0.87
Glasgow 59 0.86

I want to add column 'n' which contains new random values using the Mean and SD column values. I've done this before using:

df['n'] = np.random.normal(df['Mean'], df['SD'])

I then want to add a second column which generates a quintile rank based on the value in 'n'. I've done this using:

df['q'] = pd.qcut(df['n'], 5, labels = False)

City Mean SD n q n+1 q+1
Newcastle 60 0.81 57 5 55 5
Liverpool 62 0.91 61 1 57 4
Cardiff 65 0.87 60 1 61 1
Glasgow 59 0.86 55 3 58 3

I would like to loop these two steps to add 2000 columns, 1000 'n' columns (named 'n+1') and a 1000 'q' columns (named 'q+1').

Solution

  • This was resolved using:

    mean = df['Mean']
std_dev = df['SD']
dist = np.random.normal(mean, std_dev)

for i in range(1000):
    col_name = 'col_' + str(i)
    df[col_name] = np.random.normal(dist)'
    
    col_name_q = 'col_q_' + str(i)
    df[col_name_q] = pd.qcut(df[col_name], 5, labels = False)