How can I vectorize a loop over a Pandas DataFrame in Python?

Question

I need to make this code run fast by vectorization

final1 = pd.DataFrame()
for index, row in demo1.iterrows():
    a = np.random.choice([0, 1], size=1000, p=[1 - row['prob'], row['prob']])
    b = a * row['syb'] * (1 + row['percentage_change_syb'] / 100)
    final1 = final1.append(pd.DataFrame(b).T)

demo1 = pd.DataFrame({"syb" : [298, 388, 267, 746, 645], "prob" : [0.84, 0.46, 0.68, 0.35, 0.95], "percentage_change_syb" : [1.29, 3.45, 20, 14.9, 12.5]}) — Arnab
– Arnab, Commented Jan 8, 2019 at 9:33

Felix · Accepted Answer · 2019-01-07 08:34:34Z

1

Since you did not supply data to work against, the following code is unchecked, but should work:

def computation(prob, syb, percentage_change_syb):
    a = np.random.choice([0, 1], size=1000, p=[1 - prob, prob])
    b = a * syb * (1 + percentage_change_syb / 100)
    return b.T

final1 = computation(demo1['prob'].values, demo1['syb'].values, demo1['percentage_change_syb'].values)

For more information on the choice of operating on NumPy arrays I recommend this article.

answered Jan 7, 2019 at 8:34

Felix

1,90513 silver badges31 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Arnab Over a year ago

demo1 = pd.DataFrame({"syb" : [298, 388, 267, 746, 645], "prob" : [0.84, 0.46, 0.68, 0.35, 0.95], "percentage_change_syb" : [1.29, 3.45, 20, 14.9, 12.5]})

Arnab Over a year ago

Above is the data frame I tried your code . It is giving an error - ValueError: object too deep for desired array

Collectives™ on Stack Overflow

How can I vectorize a loop over a Pandas DataFrame in Python?

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related