python, pandas, drop rows by condition

Question

hello I need help to drop some rows by the condition: if the estimated price minus price is more than 1500 (positive) drop the row

    price      estimated price 
 0  13295                13795
 1  19990                22275
 2   7295                 6498

for example only the index 1 would be drop thank you!

sophocles · Accepted Answer · 2021-10-01 07:34:49Z

5

You can use pd.drop() in which you can drop specific rows by index. :

>>> df.drop(df[(df['estimated price']-df['price'] >= 1500)].index)

   price  estimated price
0  13295            13795
2   7295             6498

index 1 is dropped.

Note that this method assumes that your index is unique. If otherwise, boolean indexing is the better solution.

answered Sep 30, 2021 at 8:02

sophocles

13.9k3 gold badges18 silver badges37 bronze badges

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

3 Comments

Yes it only works with unique index. I assumed that though, based on the sample data we have above.

thanks for the tip, please continue pointing these things out. we are all here to learn :)

ya, only thinking sometimes about another solutions ;)

jezrael · Accepted Answer · 2021-09-30 07:52:06Z

4

Change logic for get all rows if less or equal like 1500 after subtracting in boolean indexing with Series.le:

df1 = df[df['estimated price'].sub(df['price']).le(1500)]

working like invert mask for greater like 1500:

df1 = df[~df['estimated price'].sub(df['price']).gt(1500)]

answered Sep 30, 2021 at 7:52

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges