从R到Python的By

Question

I am new to python. I will appreciate if someone can tell me how to wrap up a data table using by in python

For example

tmp2=by(data =tmp,tmp$categorie, FUN = function(x) head(x, nb))
tmp2=do.call("rbind",tmp2)

thanks in advance,

From remarks I complete my post :

 d=d[,c('categorie','titre','permanent')]
 > d
     categorie titre permanent
1          bro Vide-     false
2          bro March     false
3          bro Salon     false
4          bro Puces     false
5          bro Puces     false
6          bro March     false
2158       var   IAM     false
2159       var Valer     false
2160       zoo Aquar      true
2161       zoo Menag      true
2162       zoo Aquar      true
2163       zoo Aquar      true

> (tmp =by(data =d,d$categorie, FUN = function(x) head(x, 2)))
d$categorie: bro
  categorie titre permanent
1       bro Vide-     false
2       bro March     false
------------------------------------------------------------------------------------ 
d$categorie: var
     categorie titre permanent
2158       var   IAM     false
2159       var Valer     false
------------------------------------------------------------------------------------ 
d$categorie: zoo
     categorie titre permanent
2160       zoo Aquar      true
2161       zoo Menag      true
> 
> (tmp2=do.call("rbind",tmp))
         categorie titre permanent
bro.1          bro Vide-     false
bro.2          bro March     false
var.2158       var   IAM     false
var.2159       var Valer     false
zoo.2160       zoo Aquar      true
zoo.2161       zoo Menag      true

Answer 1

The given R code, is an example of split-apply-combine paradigm:

• split a data.frame by a the category variable
• apply a function for each group
• combine the results into a data.frame

In python , you can use pandas and groupby to do the same thing:

import pandas as pd
d = {'one' : [1., 2., 3., 4.],'two' : [1., 2., 3., 4.], 'group' : [1, 1, 2, 2]}
df = pd.DataFrame(d)
df.groupby(['group']).head(1)