Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Remove TODO sections in docs #150

Merged
merged 2 commits into from
Jan 20, 2020
Merged

Remove TODO sections in docs #150

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
411c78d
Remove TODO sections
oxinabox Jan 15, 2020
1b06c84
Move info about chainrules vs chainrulescore to FAQ
oxinabox Jan 20, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 0 additions & 1 deletion docs/make.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,6 @@ makedocs(
authors="Jarrett Revels and other contributors",
pages=[
"Introduction" => "index.md",
"Getting Started" => "getting_started.md",
"API" => "api.md",
],
)
Expand Down
27 changes: 0 additions & 27 deletions docs/src/getting_started.md
View file
Open in desktop

This file was deleted.

15 changes: 15 additions & 0 deletions docs/src/index.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -485,3 +485,18 @@ For example in `fst(x,y) = x`, then the derivative of `fst` with respect to `y`
`DoesNotExist()` represents the fact that if one perturbs the matching primal, the primal function will now error.
For example in `access(xs, n) = xs[n]` then the derivative of `access` with respect to `n` is `DoesNotExist()`.
`access([10, 20, 30], 2) = 20`, but if we add `0.1` to `2` we get `access([10, 20, 30], 2.1)` which errors as indexing can't be applied at fractional indexes.


### When to use ChainRules vs ChainRulesCore?

[ChainRulesCore.jl](https://github.com/JuliaDiff/ChainRulesCore.jl) is a light-weight dependency for defining rules for functions in your packages, without you needing to depend on ChainRules itself. It has no dependencies of its own.

[ChainRules.jl](https://github.com/JuliaDiff/ChainRules.jl) provides the full functionality, in particular it has all the rules for Base Julia and the standard libraries. Its thus a much heavier package to load.

If you only want to define rules, not use them then you probably only want to load ChainRulesCore.
AD systems making use of ChainRules should load ChainRules (rather than ChainRulesCore).

### Where should I put my rules?
In general, we recommend adding custom sensitivities to your own packages with ChainRulesCore, rather than adding them to ChainRules.jl.

A few packages currently SpecialFunctions.jl and NaNMath.jl are in ChainRules.jl but this is a short-term measure.