CODEWITHSUNDEEP
pandasdataframefindlocation
Pythn
Pythn

Reputation: 171

Finding index and column number of strings in dataframe

I have a dataframe (multiple columns) and I want to find the exact (numerical) location in the dataframe of certain strings. With location I mean I want to know both it's row number as well as it's column number (not column name).

Let's make an example with the following dataframe df:

hi      apple   fine
right   yes     okay
maybe   apple   no

I want to find the location of all apples so I want to know there is an apple on row 0 column 1 and an apple on row 2 column 1.

How can I find these numbers?

I started with looking for all places where the condition is fulfilled that that element contains indeed the string I'm looking for, with:

df[df=='apple']

This is fine but how do I get the location of all True? I tried this df.index[df=='apple'].tolist() but it gives an error because there are multiple columns. I understand I can solve this by running a for loop over all columns but wondering if there is something more straightforward. Also str.find() doesn't give me the column number. I also looked into get_loc but this only seemed to work for indices.

Upvotes: 1

Views: 706

Answers (1)

Andrej Kesely
Andrej Kesely

Reputation: 195543

You can use np.argwhere:

print(np.argwhere(df.values == "apple"))

Prints:

[[0 1]
 [2 1]]

As a list:

print(np.argwhere(df.values == "apple").tolist())

Prints:

[[0, 1], [2, 1]]

Upvotes: 2

Related Questions