How to create a function that passes pandas columns into it?

Question

I am importing a csv that looks like this. The possible values are $ and net

Type1,Type2,Type3 $,, $,net, net,$,$

I currently have it in a pandas dataframe and am wondering how I can map $ and net to different values.

I want to map $ -> DEPOSIT and net to -> NET DEP

I am passing the column into a function

def directDeposit(var):
    for value in var: 
        var.astype(str)
        if value == '$':
            var = value.map({'$':'DEPOSIT'})

df['type1'] = directDeposit(df['type1'])`

Because there are three possible types from the csv I wanted to be able to pass the type I was testing into the function so that I could repeat it for all 3. I'm not really sure where to go from here.

Enrique Ortiz Casillas · Accepted Answer · 2020-01-27T18:07:46.357

1

You can just do this:

df['type1'] = df['type1'].map({'net': 'NET DEP', '$': 'DEPOSIT', '': "No value assigned"})

When you create a function that takes a Pandas Series as an input, you don't have to loop over it. Pandas will do that for you.

edited Jan 27 '20 at 18:07

answered Jan 27 '20 at 17:53

Enrique Ortiz Casillas

519
3
10

If i use the single line of code, is there a way to allow blanks to still be passed instead of defaulting to "NET DEP" if there is a blank – Logan0015 Jan 27 '20 at 17:59
Then I get this value error. ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all(). – Logan0015 Jan 27 '20 at 18:07
Look at my new response. Should work fine. – Enrique Ortiz Casillas Jan 27 '20 at 18:08
1

Since you are keen on using map, Then it's `df['type1'].map({'net': 'NET DEP', '$': 'DEPOSIT'}).fillna(df['type1'])` – Erfan Jan 27 '20 at 18:15

score 1 · Answer 2 · answered Jan 27 '20 at 17:55

1

You can use the replace method:

df.replace({'$': 'DEPOSIT', 'net': 'NET DEP'})

answered Jan 27 '20 at 17:55

maxi.marufo

411
2
11

How to create a function that passes pandas columns into it?

2 Answers2