CODEWITHSUNDEEP
pythonpandasgroup-bypandas-groupby
Aanwar
Aanwar

Reputation: 51

Grouping by a Specific Column in the dataframe in Descending Order Pandas

so I have a data frame with 187 columns. The last column(called target) is in 1's and 0's indicating True or False. I would like to group by all the rows accordingly to the last column if it has 1 and all the rows to 0 if they equal out to 0 in the last column. 1 being shown on top and 0 after the 1's. Also, would it be possible to take out the index value of the last "1" in the last column?

So basically this:

Apologies if this is a bit confusing or if it has already been answered but I can't seem to find anything. There are about 18000 rows with approx. 14000 values equal to 1's and the rest of the 4000 equal to 0's. Would appreciate some sort of guidance here.

Upvotes: 1

Views: 44

Answers (1)

jezrael
jezrael

Reputation: 862511

First sorting values by mergesort by column target and get last row by target by DataFrame.drop_duplicates:

df1 = df.sort_values('target', ascending=False, kind='mergesort')

last = df1.drop_duplicates('target', keep='last')

If need grouping:

for name, g in df1.groupby('target'):
    print (name)
    print (g)

Upvotes: 1

Related Questions