Using : for multiple slicing in list or numpy array

Question

I'm having some difficulty trying to figure out how to do extract multiple values in a list that are spaced some indices apart. For example, given a list l = [0,1,2,3,4,5,6,7,8,9,10], I want to only extract the values [1,2,3] and [6,7,8,9]. I could do l[1:4]+l[6:-1], but is there a way such to write l[1:4,6:-1]?

This is really a ghost problem to the actual problem I am having in a pandas dataframe. I have a dataframe, df, with columns ['A','B','C','I1','D','E','F','I2','I3'], and I only want to keep the important columns ['I1', 'I2', 'I3']. Now, the current approach I am doing is

df.drop(df.columns[0:3], axis=1, inplace=True)
df.drop(df.columns[4:7], axis=1, inplace=True)

Is there a way to do it such that we can do it in 1 line without writing the column values out explicitly?

Thank you!
PS. I know pandas dataframes use numpy, and I haven't found any workarounds in numpy either, but I think the syntax to drop columns is of the standard python list format, if that makes any sense.

EDIT: I found a way to do it for numpy but it is also 2 lines, from this question. We can do:
indices = np.hstack((np.arange(0:3), np.arange(4:7))
df.drop(df.columns[indices], axis=1, inplace=True)

However, I'm still looking for 1-line generalized methods.

jezrael · Accepted Answer · 2016-12-05 08:47:24Z

3

I think you need numpy.r_ for concanecate indices:

print (np.r_[1:4, 6:10])
[1 2 3 6 7 8 9]

answered Dec 5, 2016 at 8:47

jezrael

868k103 gold badges1.4k silver badges1.3k bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

Kai Over a year ago

Ah I see! Wow, index_tricks is full of stuff. So this can only work with numpy arrays? Built in python doesn't have anything like this?

Kai Over a year ago

Hi, the numpy.r_ function doesn't work with [1:4, 6:-1], because it doesn't have any numpy array to refer to. This means we have to use [1:4, 6:array.shape[0]-1].

jezrael Over a year ago

Yes, exactly. Another possible solution is [1:4, 6:len(array)-1]

hpaulj Over a year ago

r_ expands the slices into an array of indices. Such an array can't be used to index lists, at least not without a loop or comprehension.

jezrael Over a year ago

Yes, but then you can use iloc for selecting.

|

ettanany · Accepted Answer · 2016-12-05 09:01:33Z

0

Using list comprehension, you can do:

>>> [item for item in l[1:4] + l[6:-1]]
[1, 2, 3, 6, 7, 8, 9]

You can also use extend() like this:

>>> res = l[1:4]
>>> res.extend(l[6:-1])
>>> res
[1, 2, 3, 6, 7, 8, 9]

answered Dec 5, 2016 at 9:01

ettanany

20k9 gold badges49 silver badges64 bronze badges

Collectives™ on Stack Overflow

Using : for multiple slicing in list or numpy array

2 Answers 2

6 Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related