Closed aoanla closed 10 months ago
we need to introduce @view
and stuff
We could just expand "Arrays" into its own notebook and cover @view, deepcopy, and maybe a specialist subtype (Strided? Diagonal?)
On Mon, 6 Nov 2023, 12:38 Jerry Ling, @.***> wrote:
we need to introduce @view and stuff
— Reply to this email directly, view it on GitHub https://github.com/JuliaHEP/JuliaHEP-2023/issues/29#issuecomment-1794727855, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABNRWLJ4KQLREWYXDTSDVYTYDDK37AVCNFSM6AAAAAA67MBIY2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTOOJUG4ZDOOBVGU . You are receiving this because you authored the thread.Message ID: @.***>
There are a few things to improve for the DataFrames
higgs_ml[6, [:PRI_jet_num]]
filter
, select
, combine
)input_column => filter => output_column
Also, I would fiddle with the display parameters as the output table snapshots are just too big ATM (e.g., don't really need the training rows)
Also, re that discussion in the slack just now: minor thing, but should we consolidate on one way of writing for loops (I know I like = as a heretic, but I am happy to move to in / \in ), (also other bits of minor normalisation like that)
where was the slack discussion?
The one that you contributed to where someone was upset about there being more than one way to do a thing (and used for =, in, \in etc as an example)
Personally I like in
and $\in$ - OTOH people will see all of these things used, so mentioning that is useful
I think the concensus is =
is bad, use in
or \in
Just looking at Philippe's talk one should definitely add how to select rows from a boolean selector derived from column values
I think with #36 and #33 this is done now. Anything else missing?
So, now we've run it once, this is the issue for tracking any tweaks.
Things on my list:
1) make the NRS exercise less guided now we're not time-pressured as much