adds integral arrays. Closes JuliaImages/ImagesFiltering.jl#4 #1
Conversation
src/IntegralArrays.jl
Outdated
| getindex(A::IntegralArray, i::Int...) = A.data[i...] | ||
|
|
||
| function getindex(A::IntegralArray, i::ClosedInterval...) | ||
| _boxdiff(A, left(i[1]), left(i[2]), right(i[1]), right(i[2])) |
There was a problem hiding this comment.
We can extend this to nD later on 😄
src/IntegralArrays.jl
Outdated
| @@ -1,5 +1,78 @@ | |||
| module IntegralArrays | |||
|
|
|||
| # package code goes here | |||
| using Images, ClosedIntervals | |||
There was a problem hiding this comment.
We'll likely need to persuade him to register this package.
There was a problem hiding this comment.
Just looked over the source code. Doesn't look that friendly to writing tight loops:
- the order-switching means there's a branch on construction
- the testing for empty intervals is a lot less clean than the way UnitRanges work
I wonder if we should just start a tiny package of our own.
There was a problem hiding this comment.
Hmm...for Integral Arrays, I dont think anyone would index with a reverse ordered interval. Maybe we can just use Tuples instead?
There was a problem hiding this comment.
That's a possibility, but I think it's a little dangerous because I suspect Jeff is mulling over whether we should ditch CartesianIndex and just use tuples for multidimensional indexing. (Certainly, I've given that some serious consideration.) If that happened, then indexing an IntegralImage with a 2-tuple would suddenly become very confusing.
So I think we need a type. Given the paralysis surrounding intervals, let's try making the nth package, I bet it will suddenly become the standard.
The hardest part will be the package name. Some thoughts:
- SimpleIntervals
- LRIntervals
- OrderedIntervals
- Intervulz 😄
- IntervalSets
- EnclosedIntervals
Of all these I think I like IntervalSets the best. It suggests that no arithmetic is supported. We might have the types Closed, Open, LOpen, and ROpen eventually (but perhaps start with Closed).
(Interestingly, there's a scala package with that name: https://github.com/rklaehn/intervalset)
There was a problem hiding this comment.
I'll take a shot at it 😄. I assume this shouldn't be in JuliaImages...?
There was a problem hiding this comment.
Considering that order switching and emptiness checking is the issue, what would be the way to go? Should IntervalSets be derived from UnitRanges?
There was a problem hiding this comment.
I'm not sure where to put it. Maybe https://github.com/JuliaMath.
And yes, I'd use the code for UnitRanges as the main inspiration: I'm thinking of these as ranges without any step. Obviously, don't mimic this:
julia> 0:-5
0:-1just keep whatever the user specified.
|
Changes Unknown when pulling a2aa5bb on int-arr into * on master*. |
|
Seems good to merge? |
src/IntegralArrays.jl
Outdated
|
|
||
| linearindexing(A::IntegralArray) = Base.LinearFast() | ||
| size(A::IntegralArray) = size(A.data) | ||
| getindex(A::IntegralArray, i::Int...) = A.data[i...] |
There was a problem hiding this comment.
Should we support indexing with integers at all? What use is there for the integral itself?
If we don't have to support Int, it might be good to have it throw a helpful error message (rather than just throw MethodError).
There was a problem hiding this comment.
I guess my point is that folks might unthinkingly index with integers and assume it covers the range from the beginning to the indicated position.
There was a problem hiding this comment.
Completely agree on this point. Will change it 😄
This is my best attempt at adapting #1. What do you think?
Co-authored-by: Jake Ireland <[email protected]>
|
Another commit for @mronian! Hope you're well 😃 . |
|
Wow I just went through all my notifications 🤣 I've been on a wild ride - founded a medical devices company (www.morphlelabs.com - we make high speed scanners for microscopy slides at a very affordable price) and served as its CTO up until last month. Now, back to exploring the awesome things people are building on Github and figuring out what to do next! Hope you are doing great @timholy 😄 I was lucky to have you as my mentor during GSOC 2016 🙏. Forever grateful for that experience! |
cc @timholy 😄