Drop rows on multiple conditions in pandas dataframe

Question

My df has 3 columns

df = pd.DataFrame({"col_1": (0.0, 0.0, 1.0, 1.0, 0.0, 1.0, 1.0), 
                   "col_2": (0.0, 0.24, 1.0, 0.0, 0.22, 3.11, 0.0),
                    "col_3": ("Mon", "Tue", "Thu", "Fri", "Mon", "Tue", "Thu")})

I want to drop rows where df.col_1 is 1.0 and df.col_2 is 0.0. So, I would get:

df = pd.DataFrame({"col_1": (0.0, 0.0, 1.0, 0.0, 1.0), 
                   "col_2": (0.0, 0.24, 1.0, 0.22, 3.11),
                    "col_3": ("Mon", "Tue", "Thu", "Mon", "Tue")})

I tried:

df_new = df.drop[df[(df['col_1'] == 1.0) & (df['col_2'] == 0.0)].index]

It gives me the error:

'method' object is not subscriptable

Any idea how to solve the above problem?

Asked By: Dsh M

||

Source

Answer 1

Try to filter your df with loc. It’s so powerful!
The "~" means you want to keep those with the opposite of your condition.
The ":" means you want to keep all the columns

df = df.loc[~((df['col_1'] == 1.0) & (df['col_2'] == 0.0)),:]

Answered By: Charles R

Answer 2

drop is a method, you are calling it using [], that is why it gives you:

'method' object is not subscriptable

change to () (a normal method call) and it should work:

import pandas as pd

df = pd.DataFrame({"col_1": (0.0, 0.0, 1.0, 1.0, 0.0, 1.0, 1.0),
                   "col_2": (0.0, 0.24, 1.0, 0.0, 0.22, 3.11, 0.0),
                   "col_3": ("Mon", "Tue", "Thu", "Fri", "Mon", "Tue", "Thu")})

df_new = df.drop(df[(df['col_1'] == 1.0) & (df['col_2'] == 0.0)].index)
print(df_new)

Output

   col_1  col_2 col_3
0    0.0   0.00   Mon
1    0.0   0.24   Tue
2    1.0   1.00   Thu
4    0.0   0.22   Mon
5    1.0   3.11   Tue

Answered By: Dani Mesejo

Answer 3

You can use or (|) operator for this ,
Refer this link for it pandas: multiple conditions while indexing data frame – unexpected behavior

i.e dropping rows where both conditions are met

 df = df.loc[~((df['col_1']==1) | (df['col_2']==0))]

Answered By: Saurabh

Answer 4

Put the location of row which you want remove at “location”.

df = df.drop(['location' axix=1, inplace=True]

Answered By: aliasgar paloda

Answer 5

mask = df['Product_Code'].isin(['filter1', 'filter2', 'filter3'])
df = df[~mask]
df.head()

.isin() allows you to filter the entire dataframe based on multiple values in a series. This is the least amount of code to write, compared to other solutions that I know of.

Adding the ~ inside the column wise filter reverses the logic of isin().

Answered By: Alex Slessor

Answer 6

You can also use query here:

In [4]: df.query('~(col_1 == 1 & col_2 == 0)')
Out[4]: 
   col_1  col_2 col_3
0    0.0   0.00   Mon
1    0.0   0.24   Tue
2    1.0   1.00   Thu
4    0.0   0.22   Mon
5    1.0   3.11   Tue

With ~ the query is negated, returning those observations where the condition col_1 == 1 & col_2 == 0 does not hold.

Answered By: rachwa

Drop rows on multiple conditions in pandas dataframe

Question:

Answers: