# Unexpected behaviour Raw.pick **URL:** \ **Category:** Support & Discussions\ **Created:** [January 12, 2024, 9:53am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132 "2024-01-12T09:53:33Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![ivopascal](https://avatars.discourse-cdn.com/v4/letter/i/278dde/32.png) [@ivopascal](https://mne.discourse.group/u/ivopascal)\ **Post date:** [January 12, 2024, 9:53am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/1 "2024-01-12T09:53:33Z")
I found 2 strange behaviours surrounding Raw and Pick. The first is that Raw.plot() does not support a picks argument, even though both Epochs.plot() and Evoked.plot() do support it. This is not really a problem, but it makes using MNE a little bit harder than it needs to be. The second is that Raw.pick() returns the modified instance with the picked channels (as according to documentation), but also modifies the original. - MNE version: 1.6.0 - operating system: macOS 14 ```python from mne.channels import make_standard_montage from mne.datasets import eegbci from mne.io import concatenate_raws, read_raw_edf subject = 1 runs = [6, 10, 14] # Fetch some arbitrary EEG data raw_fnames = eegbci.load_data(subject, runs) raw = concatenate_raws([read_raw_edf(f, preload=True) for f in raw_fnames]) eegbci.standardize(raw) # set channel names raw.set_montage(montage) print(raw.ch_names) # Raw will have 64 channels picked_channels = raw.pick(["C3", "C4"]) # picked_channels has 2 channels print(picked_channels.ch_names) print(raw.ch_names) # But now raw also only has C3 and C4 ``` This is not clear in the documentation, and can cause bugs. It’s probably better if the original is not modified. If the goal is in-place operations, I think it’d be clearer to not return the modified instance. Is there a reason they work like this? If not, I’d be happy to fix these things and write a PR. ---
**Author:** ![cbrnr](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/cbrnr/32/1409_2.png) [@cbrnr](https://mne.discourse.group/u/cbrnr)\ **Post date:** [January 12, 2024, 9:56am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/2 "2024-01-12T09:56:48Z")
MNE always modifies objects in place, and for convenience it also returns the modified object so that you can use it in a pipeline: [Design philosophy — MNE 1.6.1 documentation](https://mne.tools/stable/documentation/design_philosophy.html#in-place-operation) ---
**Author:** ![ivopascal](https://avatars.discourse-cdn.com/v4/letter/i/278dde/32.png) [@ivopascal](https://mne.discourse.group/u/ivopascal)\ **Post date:** [January 12, 2024, 10:14am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/3 "2024-01-12T10:14:51Z")
I’m not a fan of returning as convenience, but we can agree to disagree. How about the other behaviour that `Raw.plot()` doesn’t support a `picks` argument, even though Evoked and Epochs do? [https://mne.tools/stable/generated/mne.io.Raw.html#mne.io.Raw.plot](https://mne.tools/stable/generated/mne.io.Raw.html#mne.io.Raw.plot) [https://mne.tools/stable/generated/mne.Epochs.html#mne.Epochs.plot](https://mne.tools/stable/generated/mne.Epochs.html#mne.Epochs.plot) [https://mne.tools/stable/generated/mne.Evoked.html#mne.Evoked.plot](https://mne.tools/stable/generated/mne.Evoked.html#mne.Evoked.plot) I think adding a picks= argument to Raw.plot might help with the mostly-unified API goal ([Design philosophy — MNE 1.6.1 documentation](https://mne.tools/stable/documentation/design_philosophy.html#mostly-unified-api)). ---
**Author:** ![cbrnr](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/cbrnr/32/1409_2.png) [@cbrnr](https://mne.discourse.group/u/cbrnr)\ **Post date:** [January 12, 2024, 10:43am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/4 "2024-01-12T10:43:20Z")
> [@ivopascal](#): > > I’m not a fan of returning as convenience, but we can agree to disagree. This was not meant as an argument, but simply to point out how it is implemented. > [@ivopascal](#): > > How about the other behaviour that `Raw.plot()` doesn’t support a `picks` argument, even though Evoked and Epochs do? Yes, sure, please open an issue in our repo. ---
**Author:** ![huji-hcnl](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/huji-hcnl/32/4163_2.png) [@huji-hcnl](https://mne.discourse.group/u/huji-hcnl)\ **Post date:** [March 2, 2025, 10:44am UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/5 "2025-03-02T10:44:14Z")
This is unexpected, mostly because many other methods in the `Raw` documentation explicitly mention being performed in-place, whereas the `pick` method doesn’t. How would one “pick” a subset of their `raw` object without overwriting the original object? ---
**Author:** ![cbrnr](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/cbrnr/32/1409_2.png) [@cbrnr](https://mne.discourse.group/u/cbrnr)\ **Post date:** [March 2, 2025, 2:14pm UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/6 "2025-03-02T14:14:08Z")
`Raw.pick()` does work in-place. If this is not the case, please file a bug report. If you’re asking how would you pick a subset of channels without modifying the original object, you would use `Raw.copy().pick()`. ---
**Author:** ![richard](https://yyz2.discourse-cdn.com/free1/user_avatar/mne.discourse.group/richard/32/15_2.png) [@richard](https://mne.discourse.group/u/richard)\ **Post date:** [March 2, 2025, 8:46pm UTC](https://mne.discourse.group/t/unexpected-behaviour-raw-pick/8132/7 "2025-03-02T20:46:33Z")