CODEWITHSUNDEEP
pythonpython-3.xpandasdataframe
rsp
rsp

Reputation: 649

Need a way to create pandas Dataframe from existing Dataframe

I have a Dataframe which I get from csv file using

df = pd.read_csv('data.csv')

I want to select some of the rows of this Dataframe and create a new Dataframe but the logic to select those rows is complex and needs to be inside a function. And this filter logic uses data from that row only, not from any other rows in the Dataframe. So how can I create a new Dataframe by using this filter function so select rows from this Dataframe?

Upvotes: 0

Views: 47

Answers (1)

wolfstter
wolfstter

Reputation: 191

why are you not just a boolean mask like

idxs = df[df['foo'] == 'bar'].index.to_list()
df_slice = df.loc[idxs].copy()

alternatively

df_slice = df.query('col1 > 2 and col2 .....').copy()

If you really need to apply a function to a row i would do it like this:

# Define your function here which gets a series as input.

def check_condition(s)
   if condition: 
        return 1
   return 0

df['matches_cond'] = df[['foo', 'bar'...]].apply(
    lambda x: check_condition(x), axis=1)

And then you can slice again using loc or query.

If you need something different please add a short example of you data and the desired output

Upvotes: 1

Related Questions