From 2974d0267ffafabc4ed6f09dbc268edb7d00132b Mon Sep 17 00:00:00 2001 From: "Documenter.jl" Date: Wed, 16 Oct 2024 05:09:22 +0000 Subject: [PATCH] build based on a97b432 --- .../dev/.documenter-siteinfo.json | 2 +- DifferentiationInterface/dev/api/index.html | 66 +++++++++---------- .../dev/dev_guide/index.html | 2 +- .../dev/explanation/advanced/index.html | 2 +- .../dev/explanation/backends/index.html | 2 +- .../dev/explanation/operators/index.html | 2 +- DifferentiationInterface/dev/index.html | 2 +- .../dev/tutorials/advanced/index.html | 52 +++++++-------- .../dev/tutorials/basic/index.html | 56 ++++++++-------- 9 files changed, 93 insertions(+), 93 deletions(-) diff --git a/DifferentiationInterface/dev/.documenter-siteinfo.json b/DifferentiationInterface/dev/.documenter-siteinfo.json index ae73068f8..61d3155d1 100644 --- a/DifferentiationInterface/dev/.documenter-siteinfo.json +++ b/DifferentiationInterface/dev/.documenter-siteinfo.json @@ -1 +1 @@ -{"documenter":{"julia_version":"1.11.0","generation_timestamp":"2024-10-15T15:00:58","documenter_version":"1.7.0"}} \ No newline at end of file +{"documenter":{"julia_version":"1.11.0","generation_timestamp":"2024-10-16T05:09:15","documenter_version":"1.7.0"}} \ No newline at end of file diff --git a/DifferentiationInterface/dev/api/index.html b/DifferentiationInterface/dev/api/index.html index 532c2a4b4..c07c98143 100644 --- a/DifferentiationInterface/dev/api/index.html +++ b/DifferentiationInterface/dev/api/index.html @@ -1,5 +1,5 @@ -API · DifferentiationInterface.jl
GitHub

API

Argument wrappers

DifferentiationInterface.Constant — Type
Constant

Concrete type of Context argument which is kept constant during differentiation.

Note that an operator can be prepared with an arbitrary value of the constant. However, same-point preparation must occur with the exact value that will be reused later.

Example

julia> using DifferentiationInterface
+API · DifferentiationInterface.jl
GitHub
API
Argument wrappers
DifferentiationInterface.Constant — Type
Constant
Concrete type of Context argument which is kept constant during differentiation.
Note that an operator can be prepared with an arbitrary value of the constant. However, same-point preparation must occur with the exact value that will be reused later.
Example
julia> using DifferentiationInterface
 
 julia> import ForwardDiff
 
@@ -13,31 +13,31 @@
 julia> gradient(f, AutoForwardDiff(), [1.0, 2.0], Constant(100))
 2-element Vector{Float64}:
  200.0
- 400.0
source
First order
Pushforward
DifferentiationInterface.prepare_pushforward — Function
prepare_pushforward(f,     backend, x, tx, [contexts...]) -> prep
-prepare_pushforward(f!, y, backend, x, tx, [contexts...]) -> prep
Create a prep object that can be given to pushforward and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.prepare_pushforward_same_point — Function
prepare_pushforward_same_point(f,     backend, x, tx, [contexts...]) -> prep_same
-prepare_pushforward_same_point(f!, y, backend, x, tx, [contexts...]) -> prep_same
Create an prep_same object that can be given to pushforward and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.pushforward — Function
pushforward(f,     [prep,] backend, x, tx, [contexts...]) -> ty
-pushforward(f!, y, [prep,] backend, x, tx, [contexts...]) -> ty
Compute the pushforward of the function f at point x with a tuple of tangents tx.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named jvp.
source
DifferentiationInterface.pushforward! — Function
pushforward!(f,     dy, [prep,] backend, x, tx, [contexts...]) -> ty
-pushforward!(f!, y, dy, [prep,] backend, x, tx, [contexts...]) -> ty
Compute the pushforward of the function f at point x with a tuple of tangents tx, overwriting ty.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named jvp!.
source
DifferentiationInterface.value_and_pushforward — Function
value_and_pushforward(f,     [prep,] backend, x, tx, [contexts...]) -> (y, ty)
-value_and_pushforward(f!, y, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
Compute the value and the pushforward of the function f at point x with a tuple of tangents tx.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named value_and_jvp.
Info
Required primitive for forward mode backends.
source
DifferentiationInterface.value_and_pushforward! — Function
value_and_pushforward!(f,     dy, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
-value_and_pushforward!(f!, y, dy, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
Compute the value and the pushforward of the function f at point x with a tuple of tangents tx, overwriting ty.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named value_and_jvp!.
source
Pullback
DifferentiationInterface.prepare_pullback — Function
prepare_pullback(f,     backend, x, ty, [contexts...]) -> prep
-prepare_pullback(f!, y, backend, x, ty, [contexts...]) -> prep
Create a prep object that can be given to pullback and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.prepare_pullback_same_point — Function
prepare_pullback_same_point(f,     backend, x, ty, [contexts...]) -> prep_same
-prepare_pullback_same_point(f!, y, backend, x, ty, [contexts...]) -> prep_same
Create an prep_same object that can be given to pullback and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.pullback — Function
pullback(f,     [prep,] backend, x, ty, [contexts...]) -> tx
-pullback(f!, y, [prep,] backend, x, ty, [contexts...]) -> tx
Compute the pullback of the function f at point x with a tuple of tangents ty.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named vjp.
source
DifferentiationInterface.pullback! — Function
pullback!(f,     dx, [prep,] backend, x, ty, [contexts...]) -> tx
-pullback!(f!, y, dx, [prep,] backend, x, ty, [contexts...]) -> tx
Compute the pullback of the function f at point x with a tuple of tangents ty, overwriting dx.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named vjp!.
source
DifferentiationInterface.value_and_pullback — Function
value_and_pullback(f,     [prep,] backend, x, ty, [contexts...]) -> (y, tx)
-value_and_pullback(f!, y, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
Compute the value and the pullback of the function f at point x with a tuple of tangents ty.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named value_and_vjp.
Info
Required primitive for reverse mode backends.
source
DifferentiationInterface.value_and_pullback! — Function
value_and_pullback!(f,     dx, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
-value_and_pullback!(f!, y, dx, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
Compute the value and the pullback of the function f at point x with a tuple of tangents ty, overwriting dx.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named value_and_vjp!.
source
Derivative
DifferentiationInterface.prepare_derivative — Function
prepare_derivative(f,     backend, x, [contexts...]) -> prep
-prepare_derivative(f!, y, backend, x, [contexts...]) -> prep
Create a prep object that can be given to derivative and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.derivative — Function
derivative(f,     [prep,] backend, x, [contexts...]) -> der
-derivative(f!, y, [prep,] backend, x, [contexts...]) -> der
Compute the derivative of the function f at point x.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.derivative! — Function
derivative!(f,     der, [prep,] backend, x, [contexts...]) -> der
-derivative!(f!, y, der, [prep,] backend, x, [contexts...]) -> der
Compute the derivative of the function f at point x, overwriting der.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.value_and_derivative — Function
value_and_derivative(f,     [prep,] backend, x, [contexts...]) -> (y, der)
-value_and_derivative(f!, y, [prep,] backend, x, [contexts...]) -> (y, der)
Compute the value and the derivative of the function f at point x.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.value_and_derivative! — Function
value_and_derivative!(f,     der, [prep,] backend, x, [contexts...]) -> (y, der)
-value_and_derivative!(f!, y, der, [prep,] backend, x, [contexts...]) -> (y, der)
Compute the value and the derivative of the function f at point x, overwriting der.
To improve performance via operator preparation, refer to prepare_derivative.
source
Gradient
DifferentiationInterface.prepare_gradient — Function
prepare_gradient(f, backend, x, [contexts...]) -> prep
Create a prep object that can be given to gradient and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Jacobian
DifferentiationInterface.prepare_jacobian — Function
prepare_jacobian(f,     backend, x, [contexts...]) -> prep
-prepare_jacobian(f!, y, backend, x, [contexts...]) -> prep
Create a prep object that can be given to jacobian and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.jacobian — Function
jacobian(f,     [prep,] backend, x, [contexts...]) -> jac
-jacobian(f!, y, [prep,] backend, x, [contexts...]) -> jac
Compute the Jacobian matrix of the function f at point x.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.jacobian! — Function
jacobian!(f,     jac, [prep,] backend, x, [contexts...]) -> jac
-jacobian!(f!, y, jac, [prep,] backend, x, [contexts...]) -> jac
Compute the Jacobian matrix of the function f at point x, overwriting jac.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.value_and_jacobian — Function
value_and_jacobian(f,     [prep,] backend, x, [contexts...]) -> (y, jac)
-value_and_jacobian(f!, y, [prep,] backend, x, [contexts...]) -> (y, jac)
Compute the value and the Jacobian matrix of the function f at point x.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.value_and_jacobian! — Function
value_and_jacobian!(f,     jac, [prep,] backend, x, [contexts...]) -> (y, jac)
-value_and_jacobian!(f!, y, jac, [prep,] backend, x, [contexts...]) -> (y, jac)
Compute the value and the Jacobian matrix of the function f at point x, overwriting jac.
To improve performance via operator preparation, refer to prepare_jacobian.
source
Second order
DifferentiationInterface.SecondOrder — Type
SecondOrder
Combination of two backends for second-order differentiation.
Danger
SecondOrder backends do not support first-order operators.
Constructor
SecondOrder(outer_backend, inner_backend)
Fields
  • outer::AbstractADType: backend for the outer differentiation
  • inner::AbstractADType: backend for the inner differentiation
source
Second derivative
Hessian-vector product
DifferentiationInterface.prepare_hvp — Function
prepare_hvp(f, backend, x, tx, [contexts...]) -> prep
Create a prep object that can be given to hvp and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
DifferentiationInterface.prepare_hvp_same_point — Function
prepare_hvp_same_point(f, backend, x, tx, [contexts...]) -> prep_same
Create an prep_same object that can be given to hvp and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Hessian
DifferentiationInterface.prepare_hessian — Function
prepare_hessian(f, backend, x, [contexts...]) -> prep
Create a prep object that can be given to hessian and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Utilities
Backend queries
DifferentiationInterface.outer — Function
outer(backend::SecondOrder)
-outer(backend::AbstractADType)
Return the outer backend of a SecondOrder object, tasked with differentiation at the second order.
For any other backend type, this function acts like the identity.
source
DifferentiationInterface.inner — Function
inner(backend::SecondOrder)
-inner(backend::AbstractADType)
Return the inner backend of a SecondOrder object, tasked with differentiation at the first order.
For any other backend type, this function acts like the identity.
source
Backend switch
DifferentiationInterface.DifferentiateWith — Type
DifferentiateWith
Function wrapper that enforces differentiation with a "substitute" AD backend, possible different from the "true" AD backend that is called.
For instance, suppose a function f is not differentiable with Zygote because it involves mutation, but you know that it is differentiable with Enzyme. Then f2 = DifferentiateWith(f, AutoEnzyme()) is a new function that behaves like f, except that f2 is differentiable with Zygote (thanks to a chain rule which calls Enzyme under the hood). Moreover, any larger algorithm alg that calls f2 instead of f will also be differentiable with Zygote (as long as f was the only Zygote blocker).
Tip
This is mainly relevant for package developers who want to produce differentiable code at low cost, without writing the differentiation rules themselves. If you sprinkle a few DifferentiateWith in places where some AD backends may struggle, end users can pick from a wider variety of packages to differentiate your algorithms.
Warning
DifferentiateWith only supports out-of-place functions y = f(x) without additional context arguments. It only makes these functions differentiable if the true backend is either ForwardDiff or compatible with ChainRules. For any other true backend, the differentiation behavior is not altered by DifferentiateWith (it becomes a transparent wrapper).
Fields
  • f: the function in question, with signature f(x)
  • backend::AbstractADType: the substitute backend to use for differentiation
Note
For the substitute AD backend to be called under the hood, its package needs to be loaded in addition to the package of the true AD backend.
Constructor
DifferentiateWith(f, backend)
Example
julia> using DifferentiationInterface
+ 400.0
source
First order
Pushforward
DifferentiationInterface.prepare_pushforward — Function
prepare_pushforward(f,     backend, x, tx, [contexts...]) -> prep
+prepare_pushforward(f!, y, backend, x, tx, [contexts...]) -> prep
Create a prep object that can be given to pushforward and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.prepare_pushforward_same_point — Function
prepare_pushforward_same_point(f,     backend, x, tx, [contexts...]) -> prep_same
+prepare_pushforward_same_point(f!, y, backend, x, tx, [contexts...]) -> prep_same
Create an prep_same object that can be given to pushforward and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.pushforward — Function
pushforward(f,     [prep,] backend, x, tx, [contexts...]) -> ty
+pushforward(f!, y, [prep,] backend, x, tx, [contexts...]) -> ty
Compute the pushforward of the function f at point x with a tuple of tangents tx.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named jvp.
source
DifferentiationInterface.pushforward! — Function
pushforward!(f,     dy, [prep,] backend, x, tx, [contexts...]) -> ty
+pushforward!(f!, y, dy, [prep,] backend, x, tx, [contexts...]) -> ty
Compute the pushforward of the function f at point x with a tuple of tangents tx, overwriting ty.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named jvp!.
source
DifferentiationInterface.value_and_pushforward — Function
value_and_pushforward(f,     [prep,] backend, x, tx, [contexts...]) -> (y, ty)
+value_and_pushforward(f!, y, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
Compute the value and the pushforward of the function f at point x with a tuple of tangents tx.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named value_and_jvp.
Info
Required primitive for forward mode backends.
source
DifferentiationInterface.value_and_pushforward! — Function
value_and_pushforward!(f,     dy, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
+value_and_pushforward!(f!, y, dy, [prep,] backend, x, tx, [contexts...]) -> (y, ty)
Compute the value and the pushforward of the function f at point x with a tuple of tangents tx, overwriting ty.
To improve performance via operator preparation, refer to prepare_pushforward and prepare_pushforward_same_point.
Tip
Pushforwards are also commonly called Jacobian-vector products or JVPs. This function could have been named value_and_jvp!.
source
Pullback
DifferentiationInterface.prepare_pullback — Function
prepare_pullback(f,     backend, x, ty, [contexts...]) -> prep
+prepare_pullback(f!, y, backend, x, ty, [contexts...]) -> prep
Create a prep object that can be given to pullback and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.prepare_pullback_same_point — Function
prepare_pullback_same_point(f,     backend, x, ty, [contexts...]) -> prep_same
+prepare_pullback_same_point(f!, y, backend, x, ty, [contexts...]) -> prep_same
Create an prep_same object that can be given to pullback and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.pullback — Function
pullback(f,     [prep,] backend, x, ty, [contexts...]) -> tx
+pullback(f!, y, [prep,] backend, x, ty, [contexts...]) -> tx
Compute the pullback of the function f at point x with a tuple of tangents ty.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named vjp.
source
DifferentiationInterface.pullback! — Function
pullback!(f,     dx, [prep,] backend, x, ty, [contexts...]) -> tx
+pullback!(f!, y, dx, [prep,] backend, x, ty, [contexts...]) -> tx
Compute the pullback of the function f at point x with a tuple of tangents ty, overwriting dx.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named vjp!.
source
DifferentiationInterface.value_and_pullback — Function
value_and_pullback(f,     [prep,] backend, x, ty, [contexts...]) -> (y, tx)
+value_and_pullback(f!, y, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
Compute the value and the pullback of the function f at point x with a tuple of tangents ty.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named value_and_vjp.
Info
Required primitive for reverse mode backends.
source
DifferentiationInterface.value_and_pullback! — Function
value_and_pullback!(f,     dx, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
+value_and_pullback!(f!, y, dx, [prep,] backend, x, ty, [contexts...]) -> (y, tx)
Compute the value and the pullback of the function f at point x with a tuple of tangents ty, overwriting dx.
To improve performance via operator preparation, refer to prepare_pullback and prepare_pullback_same_point.
Tip
Pullbacks are also commonly called vector-Jacobian products or VJPs. This function could have been named value_and_vjp!.
source
Derivative
DifferentiationInterface.prepare_derivative — Function
prepare_derivative(f,     backend, x, [contexts...]) -> prep
+prepare_derivative(f!, y, backend, x, [contexts...]) -> prep
Create a prep object that can be given to derivative and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.derivative — Function
derivative(f,     [prep,] backend, x, [contexts...]) -> der
+derivative(f!, y, [prep,] backend, x, [contexts...]) -> der
Compute the derivative of the function f at point x.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.derivative! — Function
derivative!(f,     der, [prep,] backend, x, [contexts...]) -> der
+derivative!(f!, y, der, [prep,] backend, x, [contexts...]) -> der
Compute the derivative of the function f at point x, overwriting der.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.value_and_derivative — Function
value_and_derivative(f,     [prep,] backend, x, [contexts...]) -> (y, der)
+value_and_derivative(f!, y, [prep,] backend, x, [contexts...]) -> (y, der)
Compute the value and the derivative of the function f at point x.
To improve performance via operator preparation, refer to prepare_derivative.
source
DifferentiationInterface.value_and_derivative! — Function
value_and_derivative!(f,     der, [prep,] backend, x, [contexts...]) -> (y, der)
+value_and_derivative!(f!, y, der, [prep,] backend, x, [contexts...]) -> (y, der)
Compute the value and the derivative of the function f at point x, overwriting der.
To improve performance via operator preparation, refer to prepare_derivative.
source
Gradient
DifferentiationInterface.prepare_gradient — Function
prepare_gradient(f, backend, x, [contexts...]) -> prep
Create a prep object that can be given to gradient and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Jacobian
DifferentiationInterface.prepare_jacobian — Function
prepare_jacobian(f,     backend, x, [contexts...]) -> prep
+prepare_jacobian(f!, y, backend, x, [contexts...]) -> prep
Create a prep object that can be given to jacobian and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again. For in-place functions, y is mutated by f! during preparation.
source
DifferentiationInterface.jacobian — Function
jacobian(f,     [prep,] backend, x, [contexts...]) -> jac
+jacobian(f!, y, [prep,] backend, x, [contexts...]) -> jac
Compute the Jacobian matrix of the function f at point x.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.jacobian! — Function
jacobian!(f,     jac, [prep,] backend, x, [contexts...]) -> jac
+jacobian!(f!, y, jac, [prep,] backend, x, [contexts...]) -> jac
Compute the Jacobian matrix of the function f at point x, overwriting jac.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.value_and_jacobian — Function
value_and_jacobian(f,     [prep,] backend, x, [contexts...]) -> (y, jac)
+value_and_jacobian(f!, y, [prep,] backend, x, [contexts...]) -> (y, jac)
Compute the value and the Jacobian matrix of the function f at point x.
To improve performance via operator preparation, refer to prepare_jacobian.
source
DifferentiationInterface.value_and_jacobian! — Function
value_and_jacobian!(f,     jac, [prep,] backend, x, [contexts...]) -> (y, jac)
+value_and_jacobian!(f!, y, jac, [prep,] backend, x, [contexts...]) -> (y, jac)
Compute the value and the Jacobian matrix of the function f at point x, overwriting jac.
To improve performance via operator preparation, refer to prepare_jacobian.
source
Second order
DifferentiationInterface.SecondOrder — Type
SecondOrder
Combination of two backends for second-order differentiation.
Danger
SecondOrder backends do not support first-order operators.
Constructor
SecondOrder(outer_backend, inner_backend)
Fields
  • outer::AbstractADType: backend for the outer differentiation
  • inner::AbstractADType: backend for the inner differentiation
source
Second derivative
Hessian-vector product
DifferentiationInterface.prepare_hvp — Function
prepare_hvp(f, backend, x, tx, [contexts...]) -> prep
Create a prep object that can be given to hvp and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
DifferentiationInterface.prepare_hvp_same_point — Function
prepare_hvp_same_point(f, backend, x, tx, [contexts...]) -> prep_same
Create an prep_same object that can be given to hvp and its variants if they are applied at the same point x and with the same contexts.
Warning
If the function or the point changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Hessian
DifferentiationInterface.prepare_hessian — Function
prepare_hessian(f, backend, x, [contexts...]) -> prep
Create a prep object that can be given to hessian and its variants.
Warning
If the function changes in any way, the result of preparation will be invalidated, and you will need to run it again.
source
Utilities
Backend queries
DifferentiationInterface.outer — Function
outer(backend::SecondOrder)
+outer(backend::AbstractADType)
Return the outer backend of a SecondOrder object, tasked with differentiation at the second order.
For any other backend type, this function acts like the identity.
source
DifferentiationInterface.inner — Function
inner(backend::SecondOrder)
+inner(backend::AbstractADType)
Return the inner backend of a SecondOrder object, tasked with differentiation at the first order.
For any other backend type, this function acts like the identity.
source
Backend switch
DifferentiationInterface.DifferentiateWith — Type
DifferentiateWith
Function wrapper that enforces differentiation with a "substitute" AD backend, possible different from the "true" AD backend that is called.
For instance, suppose a function f is not differentiable with Zygote because it involves mutation, but you know that it is differentiable with Enzyme. Then f2 = DifferentiateWith(f, AutoEnzyme()) is a new function that behaves like f, except that f2 is differentiable with Zygote (thanks to a chain rule which calls Enzyme under the hood). Moreover, any larger algorithm alg that calls f2 instead of f will also be differentiable with Zygote (as long as f was the only Zygote blocker).
Tip
This is mainly relevant for package developers who want to produce differentiable code at low cost, without writing the differentiation rules themselves. If you sprinkle a few DifferentiateWith in places where some AD backends may struggle, end users can pick from a wider variety of packages to differentiate your algorithms.
Warning
DifferentiateWith only supports out-of-place functions y = f(x) without additional context arguments. It only makes these functions differentiable if the true backend is either ForwardDiff or compatible with ChainRules. For any other true backend, the differentiation behavior is not altered by DifferentiateWith (it becomes a transparent wrapper).
Fields
  • f: the function in question, with signature f(x)
  • backend::AbstractADType: the substitute backend to use for differentiation
Note
For the substitute AD backend to be called under the hood, its package needs to be loaded in addition to the package of the true AD backend.
Constructor
DifferentiateWith(f, backend)
Example
julia> using DifferentiationInterface
 
 julia> import FiniteDiff, ForwardDiff, Zygote
 
@@ -62,7 +62,7 @@
 julia> Zygote.gradient(alg, [3.0, 5.0])[1]
 2-element Vector{Float64}:
  42.0
- 70.0
source
Sparsity detection
DifferentiationInterface.DenseSparsityDetector — Type
DenseSparsityDetector
Sparsity pattern detector satisfying the detection API of ADTypes.jl.
The nonzeros in a Jacobian or Hessian are detected by computing the relevant matrix with dense AD, and thresholding the entries with a given tolerance (which can be numerically inaccurate). This process can be very slow, and should only be used if its output can be exploited multiple times to compute many sparse matrices.
Danger
In general, the sparsity pattern you obtain can depend on the provided input x. If you want to reuse the pattern, make sure that it is input-agnostic.
Warning
DenseSparsityDetector functionality is now located in a package extension, please load the SparseArrays.jl standard library before you use it.
Fields
  • backend::AbstractADType is the dense AD backend used under the hood
  • atol::Float64 is the minimum magnitude of a matrix entry to be considered nonzero
Constructor
DenseSparsityDetector(backend; atol, method=:iterative)
The keyword argument method::Symbol can be either:
  • :iterative: compute the matrix in a sequence of matrix-vector products (memory-efficient)
  • :direct: compute the matrix all at once (memory-hungry but sometimes faster).
Note that the constructor is type-unstable because method ends up being a type parameter of the DenseSparsityDetector object (this is not part of the API and might change).
Examples
using ADTypes, DifferentiationInterface, SparseArrays
+ 70.0
source
Sparsity detection
DifferentiationInterface.DenseSparsityDetector — Type
DenseSparsityDetector
Sparsity pattern detector satisfying the detection API of ADTypes.jl.
The nonzeros in a Jacobian or Hessian are detected by computing the relevant matrix with dense AD, and thresholding the entries with a given tolerance (which can be numerically inaccurate). This process can be very slow, and should only be used if its output can be exploited multiple times to compute many sparse matrices.
Danger
In general, the sparsity pattern you obtain can depend on the provided input x. If you want to reuse the pattern, make sure that it is input-agnostic.
Warning
DenseSparsityDetector functionality is now located in a package extension, please load the SparseArrays.jl standard library before you use it.
Fields
  • backend::AbstractADType is the dense AD backend used under the hood
  • atol::Float64 is the minimum magnitude of a matrix entry to be considered nonzero
Constructor
DenseSparsityDetector(backend; atol, method=:iterative)
The keyword argument method::Symbol can be either:
  • :iterative: compute the matrix in a sequence of matrix-vector products (memory-efficient)
  • :direct: compute the matrix all at once (memory-hungry but sometimes faster).
Note that the constructor is type-unstable because method ends up being a type parameter of the DenseSparsityDetector object (this is not part of the API and might change).
Examples
using ADTypes, DifferentiationInterface, SparseArrays
 import ForwardDiff
 
 detector = DenseSparsityDetector(AutoForwardDiff(); atol=1e-5, method=:direct)
@@ -85,13 +85,13 @@
 # output
 
 1×2 SparseMatrixCSC{Bool, Int64} with 1 stored entry:
- 1  ⋅
source
Internals
The following is not part of the public API.
DifferentiationInterface.BatchSizeSettings — Type
BatchSizeSettings{B,singlebatch,aligned}
Configuration for the batch size deduced from a backend and a sample array of length N.
Type parameters
  • B::Int: batch size
  • singlebatch::Bool: whether B > N
  • aligned::Bool: whether N % B == 0
Fields
  • N::Int: array length
  • A::Int: number of batches A = div(N, B, RoundUp)
  • B_last::Int: size of the last batch (if aligned is false)
source
ADTypes.mode — Method
mode(backend::SecondOrder)
Return the outer mode of the second-order backend.
source
DifferentiationInterface.basis — Method
basis(backend, a::AbstractArray, i)
Construct the i-th standard basis array in the vector space of a with element type eltype(a).
Note
If an AD backend benefits from a more specialized basis array implementation, this function can be extended on the backend type.
source
DifferentiationInterface.multibasis — Method
multibasis(backend, a::AbstractArray, inds::AbstractVector)
Construct the sum of the i-th standard basis arrays in the vector space of a with element type eltype(a), for all i ∈ inds.
Note
If an AD backend benefits from a more specialized basis array implementation, this function can be extended on the backend type.
source
DifferentiationInterface.prepare!_derivative — Function
prepare!_derivative(f,     prep, backend, x, [contexts...]) -> new_prep
-prepare!_derivative(f!, y, prep, backend, x, [contexts...]) -> new_prep
Same behavior as prepare_derivative but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_gradient — Function
prepare!_gradient(f, prep, backend, x, [contexts...]) -> new_prep
Same behavior as prepare_gradient but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_hessian — Function
prepare!_hessian(f, backend, x, [contexts...]) -> new_prep
Same behavior as prepare_hessian but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_hvp — Function
prepare!_hvp(f, backend, x, tx, [contexts...]) -> new_prep
Same behavior as prepare_hvp but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_jacobian — Function
prepare!_jacobian(f,     prep, backend, x, [contexts...]) -> new_prep
-prepare!_jacobian(f!, y, prep, backend, x, [contexts...]) -> new_prep
Same behavior as prepare_jacobian but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_pullback — Function
prepare!_pullback(f,     prep, backend, x, ty, [contexts...]) -> new_prep
-prepare!_pullback(f!, y, prep, backend, x, ty, [contexts...]) -> new_prep
Same behavior as prepare_pullback but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source
DifferentiationInterface.prepare!_pushforward — Function
prepare!_pushforward(f,     prep, backend, x, tx, [contexts...]) -> new_prep
-prepare!_pushforward(f!, y, prep, backend, x, tx, [contexts...]) -> new_prep
Same behavior as prepare_pushforward but can modify an existing prep object to avoid some allocations.
There is no guarantee that prep will be mutated, or that performance will be improved compared to preparation from scratch.
Danger
For efficiency, this function needs to rely on backend package internals, therefore it not protected by semantic versioning.
source

+
diff --git a/DifferentiationInterface/dev/dev_guide/index.html b/DifferentiationInterface/dev/dev_guide/index.html index 5ef3a354e..21e78d94a 100644 --- a/DifferentiationInterface/dev/dev_guide/index.html +++ b/DifferentiationInterface/dev/dev_guide/index.html @@ -4,4 +4,4 @@ startOnLoad: true, theme: "neutral" }); - + diff --git a/DifferentiationInterface/dev/explanation/advanced/index.html b/DifferentiationInterface/dev/explanation/advanced/index.html index 52777973b..18eff5b55 100644 --- a/DifferentiationInterface/dev/explanation/advanced/index.html +++ b/DifferentiationInterface/dev/explanation/advanced/index.html @@ -4,4 +4,4 @@ startOnLoad: true, theme: "neutral" }); - + diff --git a/DifferentiationInterface/dev/explanation/backends/index.html b/DifferentiationInterface/dev/explanation/backends/index.html index fbf64c567..c0bc64788 100644 --- a/DifferentiationInterface/dev/explanation/backends/index.html +++ b/DifferentiationInterface/dev/explanation/backends/index.html @@ -4,4 +4,4 @@ startOnLoad: true, theme: "neutral" }); - + diff --git a/DifferentiationInterface/dev/explanation/operators/index.html b/DifferentiationInterface/dev/explanation/operators/index.html index dae97fe93..eb55e9962 100644 --- a/DifferentiationInterface/dev/explanation/operators/index.html +++ b/DifferentiationInterface/dev/explanation/operators/index.html @@ -5,4 +5,4 @@ startOnLoad: true, theme: "neutral" }); - + diff --git a/DifferentiationInterface/dev/index.html b/DifferentiationInterface/dev/index.html index 230f4a7c6..39a6ca785 100644 --- a/DifferentiationInterface/dev/index.html +++ b/DifferentiationInterface/dev/index.html @@ -20,4 +20,4 @@ startOnLoad: true, theme: "neutral" }); - + diff --git a/DifferentiationInterface/dev/tutorials/advanced/index.html b/DifferentiationInterface/dev/tutorials/advanced/index.html index 1e98c34ff..e36776e0e 100644 --- a/DifferentiationInterface/dev/tutorials/advanced/index.html +++ b/DifferentiationInterface/dev/tutorials/advanced/index.html @@ -88,34 +88,34 @@ [2, 4] [5, 7] [6, 8]

Sparsity speedup

When preparation is used, the speedup due to sparsity becomes very visible in large dimensions.

xbig = rand(1000)
jac_prep_dense = prepare_jacobian(f_sparse_vector, dense_first_order_backend, zero(xbig))
-@benchmark jacobian($f_sparse_vector, $jac_prep_dense, $dense_first_order_backend, $xbig)
BenchmarkTools.Trial: 409 samples with 1 evaluation.
- Range (min … max):   5.498 ms … 201.260 ms  ┊ GC (min … max):  0.00% … 97.05%
- Time  (median):      6.741 ms               ┊ GC (median):    14.51%
- Time  (mean ± σ):   12.197 ms ±  28.793 ms  ┊ GC (mean ± σ):  50.11% ± 19.50%
+@benchmark jacobian($f_sparse_vector, $jac_prep_dense, $dense_first_order_backend, $xbig)
BenchmarkTools.Trial: 420 samples with 1 evaluation.
+ Range (min … max):   4.769 ms … 171.034 ms  ┊ GC (min … max):  9.38% … 94.39%
+ Time  (median):      5.371 ms               ┊ GC (median):    12.91%
+ Time  (mean ± σ):   12.087 ms ±  28.191 ms  ┊ GC (mean ± σ):  49.53% ± 19.28%
 
-  █▂                                                            
-  ██▅▇▁▁▁▇▅▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▄▁▄▅▅ ▆
-  5.5 ms        Histogram: log(frequency) by time       184 ms <
+  █▂                                                            
+  ██▇▇▇▆▇▆▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▄▄▁▁▁▁▁▇ ▆
+  4.77 ms       Histogram: log(frequency) by time       165 ms <
 
  Memory estimate: 57.63 MiB, allocs estimate: 1515.
jac_prep_sparse = prepare_jacobian(f_sparse_vector, sparse_first_order_backend, zero(xbig))
 @benchmark jacobian($f_sparse_vector, $jac_prep_sparse, $sparse_first_order_backend, $xbig)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
- Range (min … max):  28.112 μs …  3.628 ms  ┊ GC (min … max):  0.00% … 95.88%
- Time  (median):     33.192 μs              ┊ GC (median):     0.00%
- Time  (mean ± σ):   39.910 μs ± 90.117 μs  ┊ GC (mean ± σ):  14.38% ±  7.01%
+ Range (min … max):  23.013 μs …  4.387 ms  ┊ GC (min … max):  0.00% … 97.80%
+ Time  (median):     28.624 μs              ┊ GC (median):     0.00%
+ Time  (mean ± σ):   37.436 μs ± 84.192 μs  ┊ GC (mean ± σ):  12.97% ±  7.29%
 
-        ▂▄▆▇▇██▇▅▃▁                                            
-  ▁▁▂▃▄████████████▇▆▅▄▃▃▂▂▂▂▂▁▁▁▂▁▁▂▁▁▁▁▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁ ▃
-  28.1 μs         Histogram: frequency by time        52.7 μs <
+  ▆███▅▂▂▂▂                                            ▂▂     ▂
+  ██████████▇▅▄▅▄▃▃▁▁▁▁▁▁▃▁▁▁▁▃▁▁▃▁▁▁▁▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▅███▇▇▇ █
+  23 μs        Histogram: log(frequency) by time       145 μs <
 
  Memory estimate: 305.31 KiB, allocs estimate: 27.

Better memory use can be achieved by pre-allocating the matrix from the preparation result (so that it has the correct structure).

jac_buffer = similar(sparsity_pattern(jac_prep_sparse), eltype(xbig))
 @benchmark jacobian!($f_sparse_vector, $jac_buffer, $jac_prep_sparse, $sparse_first_order_backend, $xbig)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
- Range (min … max):  23.474 μs …  3.637 ms  ┊ GC (min … max):  0.00% … 97.72%
- Time  (median):     27.421 μs              ┊ GC (median):     0.00%
- Time  (mean ± σ):   32.961 μs ± 75.771 μs  ┊ GC (mean ± σ):  11.53% ±  5.62%
+ Range (min … max):  18.955 μs …  4.647 ms  ┊ GC (min … max):  0.00% … 99.07%
+ Time  (median):     24.085 μs              ┊ GC (median):     0.00%
+ Time  (mean ± σ):   31.704 μs ± 73.846 μs  ┊ GC (mean ± σ):  10.57% ±  6.14%
 
-  ▃▇█▇▅▃▂▁▂▂                                                  ▂
-  ████████████▇▇▇▆▇▅▅▅▄▄▃▃▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▃▁▁▃▁▃▄▆▇ █
-  23.5 μs      Histogram: log(frequency) by time       112 μs <
+  ▃▆▇█▅▂▁▁▂▂                                          ▁▂▂▁    ▂
+  ███████████▇▆▆▆▆▅▃▃▃▁▁▁▁▁▁▁▁▁▃▁▁▁▁▁▁▁▁▁▃▁▁▁▁▁▁▁▁▁▁▃▆████▇▆▇ █
+  19 μs        Histogram: log(frequency) by time       116 μs <
 
  Memory estimate: 234.75 KiB, allocs estimate: 18.

And for optimal speed, one should write non-allocating and type-stable functions.

function f_sparse_vector!(y::AbstractVector, x::AbstractVector)
     n = length(x)
@@ -130,17 +130,17 @@
 ybig ≈ f_sparse_vector(xbig)
true

In this case, the sparse Jacobian should also become non-allocating (for our specific choice of backend).

jac_prep_sparse_nonallocating = prepare_jacobian(f_sparse_vector!, zero(ybig), sparse_first_order_backend, zero(xbig))
 jac_buffer = similar(sparsity_pattern(jac_prep_sparse_nonallocating), eltype(xbig))
 @benchmark jacobian!($f_sparse_vector!, $ybig, $jac_buffer, $jac_prep_sparse_nonallocating, $sparse_first_order_backend, $xbig)
BenchmarkTools.Trial: 10000 samples with 1 evaluation.
- Range (min … max):  13.034 μs … 56.164 μs  ┊ GC (min … max): 0.00% … 0.00%
- Time  (median):     13.325 μs              ┊ GC (median):    0.00%
- Time  (mean ± σ):   13.511 μs ±  1.268 μs  ┊ GC (mean ± σ):  0.00% ± 0.00%
+ Range (min … max):  12.954 μs … 31.549 μs  ┊ GC (min … max): 0.00% … 0.00%
+ Time  (median):     13.215 μs              ┊ GC (median):    0.00%
+ Time  (mean ± σ):   13.363 μs ±  1.029 μs  ┊ GC (mean ± σ):  0.00% ± 0.00%
 
-  ▁██▃   ▁                                                    ▂
-  ████▇████▆▅▃▅▄▃▅▃▄▆▅▅▅▄▄▄▁▄▃▄▅▄▅▄▃▄▄▁▃▃▄▁▁▁▁▁▁▁▁▁▃█▆▁▁▁▃▄▆▇ █
-  13 μs        Histogram: log(frequency) by time      21.1 μs <
+   █▅                                                          
+  ▄██▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▂▂▂▂▂▁▁▁▂▂▁▁▁▂▁▁▁▂▂▂ ▂
+  13 μs           Histogram: frequency by time        21.1 μs <
 
  Memory estimate: 0 bytes, allocs estimate: 0.
+ diff --git a/DifferentiationInterface/dev/tutorials/basic/index.html b/DifferentiationInterface/dev/tutorials/basic/index.html index eb852979e..9eff34ce5 100644 --- a/DifferentiationInterface/dev/tutorials/basic/index.html +++ b/DifferentiationInterface/dev/tutorials/basic/index.html @@ -13,14 +13,14 @@ 8.0 10.0

Was that fast? BenchmarkTools.jl helps you answer that question.

using BenchmarkTools
 
-@benchmark gradient($f, $backend, $x)
BenchmarkTools.Trial: 10000 samples with 151 evaluations.
- Range (min … max):  678.543 ns … 203.603 μs  ┊ GC (min … max): 0.00% … 92.52%
- Time  (median):     992.305 ns               ┊ GC (median):    0.00%
- Time  (mean ± σ):   973.630 ns ±   3.445 μs  ┊ GC (mean ± σ):  8.47% ±  3.17%
+@benchmark gradient($f, $backend, $x)
BenchmarkTools.Trial: 10000 samples with 152 evaluations.
+ Range (min … max):  678.178 ns … 172.748 μs  ┊ GC (min … max): 0.00% … 99.35%
+ Time  (median):     987.605 ns               ┊ GC (median):    0.00%
+ Time  (mean ± σ):   955.980 ns ±   3.049 μs  ┊ GC (mean ± σ):  7.83% ±  3.53%
 
-    ▅█▃                            ▃█▄                           
-  ▆▇███▅▃▃▃▃▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▁▃▅████▅▄▃▄▄▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂ ▃
-  679 ns           Histogram: frequency by time         1.26 μs <
+   ▆█                        ▅▄                                  
+  ▄██▅▂▂▃▄▃▂▂▂▂▂▂▂▂▂▂▁▂▂▂▁▁▃▇██▇▄▃▄▄▃▃▂▂▂▂▂▂▂▂▂▂▂▂▁▂▁▂▂▂▂▂▂▂▁▂▂ ▃
+  678 ns           Histogram: frequency by time         1.39 μs <
 
  Memory estimate: 848 bytes, allocs estimate: 6.

Not bad, but you can do better.

Overwriting a gradient

Since you know how much space your gradient will occupy (the same as your input x), you can pre-allocate that memory and offer it to AD. Some backends get a speed boost from this trick.

grad = similar(x)
 gradient!(f, grad, backend, x)
@@ -29,16 +29,16 @@
   4.0
   6.0
   8.0
- 10.0

The bang indicates that one of the arguments of gradient! might be mutated. More precisely, our convention is that every positional argument between the function and the backend is mutated.

@benchmark gradient!($f, $grad, $backend, $x)
BenchmarkTools.Trial: 10000 samples with 221 evaluations.
- Range (min … max):  330.882 ns … 113.651 μs  ┊ GC (min … max): 0.00% … 99.37%
- Time  (median):     536.833 ns               ┊ GC (median):    0.00%
- Time  (mean ± σ):   510.106 ns ±   1.992 μs  ┊ GC (mean ± σ):  9.58% ±  3.33%
+ 10.0

The bang indicates that one of the arguments of gradient! might be mutated. More precisely, our convention is that every positional argument between the function and the backend is mutated.

@benchmark gradient!($f, $grad, $backend, $x)
BenchmarkTools.Trial: 10000 samples with 224 evaluations.
+ Range (min … max):  330.170 ns … 106.391 μs  ┊ GC (min … max): 0.00% … 99.44%
+ Time  (median):     531.799 ns               ┊ GC (median):    0.00%
+ Time  (mean ± σ):   493.220 ns ±   1.797 μs  ┊ GC (mean ± σ):  8.88% ±  3.75%
 
-   █▄                                          ▃▂                
-  ▄██▅▃▂▂▂▂▂▃▃▂▂▂▂▂▂▂▂▂▂▁▁▂▂▂▂▂▂▂▂▂▂▂▁▂▂▁▂▂▂▃▅███▇▄▃▂▂▂▃▃▃▃▂▂▂▂ ▃
-  331 ns           Histogram: frequency by time          617 ns <
+   █                                              ▂              
+  ▅██▃▂▂▂▂▂▂▃▃▂▂▂▂▂▂▁▂▂▂▂▂▂▁▂▂▂▂▂▂▁▁▁▁▁▁▂▂▁▁▁▁▂▂▄▆██▆▃▂▂▂▂▂▃▃▃▂ ▃
+  330 ns           Histogram: frequency by time          595 ns <
 
- Memory estimate: 528 bytes, allocs estimate: 3.

For some reason the in-place version is not much better than your first attempt. However, it makes fewer allocations, thanks to the gradient vector you provided. Don't worry, you can get even more performance.

Preparing for multiple gradients

Internally, ForwardDiff.jl creates some data structures to keep track of things. These objects can be reused between gradient computations, even on different input values. We abstract away the preparation step behind a backend-agnostic syntax:

prep = prepare_gradient(f, backend, zero(x))
DifferentiationInterfaceForwardDiffExt.ForwardDiffGradientPrep{ForwardDiff.GradientConfig{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5, Vector{ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}}}}(ForwardDiff.GradientConfig{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5, Vector{ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}}}((Partials(1.0, 0.0, 0.0, 0.0, 0.0), Partials(0.0, 1.0, 0.0, 0.0, 0.0), Partials(0.0, 0.0, 1.0, 0.0, 0.0), Partials(0.0, 0.0, 0.0, 1.0, 0.0), Partials(0.0, 0.0, 0.0, 0.0, 1.0)), ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}[Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(0.0,0.0,1.0,6.8995867595565e-310,0.0,6.8995898395981e-310), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(1.0,0.0,0.0,0.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(1.0,0.0,0.0,0.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(1.0,0.0,0.0,0.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(1.0,0.0,0.0,0.0,0.0,0.0)]))

You don't need to know what this object is, you just need to pass it to the gradient operator. Note that preparation does not depend on the actual components of the vector x, just on its type and size. You can thus reuse the prep for different values of the input.

grad = similar(x)
+ Memory estimate: 528 bytes, allocs estimate: 3.

For some reason the in-place version is not much better than your first attempt. However, it makes fewer allocations, thanks to the gradient vector you provided. Don't worry, you can get even more performance.

Preparing for multiple gradients

Internally, ForwardDiff.jl creates some data structures to keep track of things. These objects can be reused between gradient computations, even on different input values. We abstract away the preparation step behind a backend-agnostic syntax:

prep = prepare_gradient(f, backend, zero(x))
DifferentiationInterfaceForwardDiffExt.ForwardDiffGradientPrep{ForwardDiff.GradientConfig{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5, Vector{ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}}}}(ForwardDiff.GradientConfig{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5, Vector{ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}}}((Partials(1.0, 0.0, 0.0, 0.0, 0.0), Partials(0.0, 1.0, 0.0, 0.0, 0.0), Partials(0.0, 0.0, 1.0, 0.0, 0.0), Partials(0.0, 0.0, 0.0, 1.0, 0.0), Partials(0.0, 0.0, 0.0, 0.0, 1.0)), ForwardDiff.Dual{ForwardDiff.Tag{typeof(Main.f), Float64}, Float64, 5}[Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(1.0,1.0,0.0,0.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(2.0,0.0,1.0,0.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(3.0,0.0,0.0,1.0,0.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(4.0,0.0,0.0,0.0,1.0,0.0), Dual{ForwardDiff.Tag{typeof(Main.f), Float64}}(5.0,0.0,0.0,0.0,0.0,1.0)]))

You don't need to know what this object is, you just need to pass it to the gradient operator. Note that preparation does not depend on the actual components of the vector x, just on its type and size. You can thus reuse the prep for different values of the input.

grad = similar(x)
 gradient!(f, grad, prep, backend, x)
 grad  # has been mutated
5-element Vector{Float64}:
   2.0
@@ -46,13 +46,13 @@
   6.0
   8.0
  10.0

Preparation makes the gradient computation much faster, and (in this case) allocation-free.

@benchmark gradient!($f, $grad, $prep, $backend, $x)
BenchmarkTools.Trial: 10000 samples with 995 evaluations.
- Range (min … max):  32.029 ns … 51.795 ns  ┊ GC (min … max): 0.00% … 0.00%
- Time  (median):     32.401 ns              ┊ GC (median):    0.00%
- Time  (mean ± σ):   32.676 ns ±  1.458 ns  ┊ GC (mean ± σ):  0.00% ± 0.00%
+ Range (min … max):  27.509 ns … 45.099 ns  ┊ GC (min … max): 0.00% … 0.00%
+ Time  (median):     27.861 ns              ┊ GC (median):    0.00%
+ Time  (mean ± σ):   28.124 ns ±  1.281 ns  ┊ GC (mean ± σ):  0.00% ± 0.00%
 
-  ▄▇██▅▂                                                  ▁▁▁ ▂
-  ██████▇▆█▇▅▄▃▄▁▅▄▆▆▅▄▆▃▃▄▁▁▃▄▁▁▄▁▄▁▁▁▁▄▁▃▁▄▃▁▁▁▁▄▄▄▄▄▆█████ █
-  32 ns        Histogram: log(frequency) by time      39.7 ns <
+  ▅▇█▇▆▆▄▂                                                ▁▁  ▂
+  █████████▇▄▄▅▁▃▄▄▅▄▅▃▁▁▃▄▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▃▁▄▁▁▁▁▄▁▆▇████ █
+  27.5 ns      Histogram: log(frequency) by time      35.1 ns <
 
  Memory estimate: 0 bytes, allocs estimate: 0.

Beware that the prep object is nearly always mutated by differentiation operators, even though it is given as the last positional argument.

Switching backends

The whole point of DifferentiationInterface.jl is that you can easily experiment with different AD solutions. Typically, for gradients, reverse mode AD might be a better fit, so let's try Zygote.jl!

import Zygote
 
@@ -64,17 +64,17 @@
  10.0

And you can run the same benchmarks to see what you gained (although such a small input may not be realistic):

prep2 = prepare_gradient(f, backend2, zero(x))
 
 @benchmark gradient!($f, $grad, $prep2, $backend2, $x)
BenchmarkTools.Trial: 10000 samples with 993 evaluations.
- Range (min … max):  35.242 ns …  31.984 μs  ┊ GC (min … max):  0.00% … 99.66%
- Time  (median):     61.877 ns               ┊ GC (median):     0.00%
- Time  (mean ± σ):   66.542 ns ± 525.144 ns  ┊ GC (mean ± σ):  13.62% ±  1.73%
+ Range (min … max):  34.616 ns …  29.294 μs  ┊ GC (min … max):  0.00% … 99.71%
+ Time  (median):     60.203 ns               ┊ GC (median):     0.00%
+ Time  (mean ± σ):   64.124 ns ± 479.852 ns  ┊ GC (mean ± σ):  12.93% ±  1.73%
 
-   ▅█▄                           ▂▅▇▃          ▁▁▁              
-  ▅███▇▄▃▂▂▁▁▁▁▂▂▁▁▁▁▁▁▁▁▁▁▁▁▁▁▁▄████▆▃▂▁▁▁▂▃▄▇███▇▄▃▂▁▁▁▁▂▁▂▁ ▃
-  35.2 ns         Histogram: frequency by time         83.6 ns <
+   █                             ▁▁▃                            
+  ▆██▃▃▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▁▂▂▂▁▂▂▄███▇▄▃▂▂▂▂▂▃▃▄▅▆▆▅▅▃▂▂▂▂▂▂▂▂▂ ▃
+  34.6 ns         Histogram: frequency by time         81.3 ns <
 
  Memory estimate: 96 bytes, allocs estimate: 2.

In short, DifferentiationInterface.jl allows for easy testing and comparison of AD backends. If you want to go further, check out the documentation of DifferentiationInterfaceTest.jl. This related package provides benchmarking utilities to compare backends and help you select the one that is best suited for your problem.

+