[FEATURE-REQUEST]

[Leo-Ji2020]

I have a request: After groupby a vaex dataframe, how can I get the first row of each group to build a new vaex dataframe? for example:

 import numpy as np
 import vaex

 x=np.random.randint(1, 5, 20)
 df = vaex.from_arrays(x=x, y=x**2)

 dff = df.groupby(df.x)

 now, I want to concat the first row of each group to build a new dataframe. how to do that?

[kmcentush]

This will actually have a similar solution to https://github.com/vaexio/vaex/issues/1448, so I recommend you follow that issue in addition to this one.

[JovanVeljanoski]

Indeed, like @kmcentush said.

Anyway here is a belated example

import vaex

df = vaex.example()
g = df.groupby('id')

dfs = []
for _, df_tmp in g:
    dfs.append(df_tmp[:1]) 

df2 = vaex.concat(dfs)
df2 = df2.trim()  # this will make things faster.