I need to create a new column in a csv called BTTS, which is based on two other columns, FTHG and FTAG. If FTHG & FTAG are both greater than zero, BTTS should be 1. Otherwise it should be zero.
What's the best way to do this in pandas / numpys?
2 Answers
I'm not sure, what the best way is. But here is one solution using pandas loc method:
df.loc[((df['FTHG'] > 0) & (df['FTAG'] > 0)),'BTTS'] = 1
df['BTTS'].fillna(0, inplace=True)
Another solution using pandas apply method:
def check_greater_zero(row):
return 1 if row['FTHG'] > 0 & row['FTAG'] > 0 else 0
df['BTTS'] = df.apply(check_greater_zero, axis=1)
EDIT:
As stated in the comments, the first, vectorized, implementation is more efficient.
2 Comments
gold_cy
your answer contains a fallacy,
apply is not more efficient, in fact it is less efficient because it is the same as looping over the dataframe. you are better off using a vectorized approach like you have listed in your first snippet
Jannik
You are right! I was testing my code with just a small example in which case the second approach was more efficient than the first approach. But as it was just a very small example I agree: The bigger the DataFrame, the more efficient the first approach is.
I dont know if this is the best way to do it but this works :)
df['BTTS'] = [1 if x == y == 1 else 0 for x, y in zip(df['FTAG'], df['FTHG'])]
