Extend basic operations #174
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This is predicated on a new unary-operator function that replicates the binary-operator, but with a single argument.
... and (NUMBER, TENSOR) arguments This simplifes the workload if, e.g., adding a single number to a tensor. No need to create a tensor all of a single value just to apply the operation.
stylewarning
requested changes
Jul 3, 2022
src/high-level/abstract-tensor.lisp
Outdated
| @@ -245,6 +279,51 @@ If TARGET is not specified then a new tensor is created with the same element ty | |||
| ;; choice, like integers. | |||
| (binary-operator #'expt source1 source2 target))) | |||
|
|
|||
| (define-backend-function .max (source1 source2 &optional target) | |||
There was a problem hiding this comment.
Can you use DEFINE-EXTENSIBLE-FUNCTION without specialization here so that one has the option to micro-optimize on tensor type in the future?
Tests are for: .+ .- .* ./ .max .min .exp .log
|
Addressed both comments @stylewarning I'd read the Developer How-To umpteen times and missed that DEFINE-EXTENSIBLE-FUNCTION could be used without specialization. 🤦 |
stylewarning
approved these changes
Jul 4, 2022
stylewarning
pushed a commit
that referenced
this pull request
Jul 4, 2022
|
Great, thanks! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Extend basic basic operations (i.e. elementwise ops) in three ways:
numberinstead of anabstract-tensortype as one of the arguments(.+ A 10)behaves as it reads - i.e. add 10 to each element ofA. An alternative would have been to create a tensor the same type/shape asAand set all elements to 10...and then...add toA.maxand.min(.max 0.0 A).expand.log(natural log)unary-operatorfunction, which will look suspiciously like the existingbinary-operator, but with only one argumentAdditional thoughts
.maxand.minfunctions suffer from the existing issue that if the type of the arguments differ then we can get a TYPE-ERROR from the SETF. See issue #173. So rather than(.max A 0), it's better to be explicit and write(.max A 0.0).