CRFMNES

https://docs.rs/crfmnes/1.0.0/crfmnes/
Admin
CRFMNES is a derivative-free optimizer for high-dimensional black-box problems, using covariance...
Tokens:9,089
Snippets:120
Trust Score:9.1
Update:5 months ago
Show doc for...
Tokens:
Raw
### Get Sample Vectors - Trial::x

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Retrieves a matrix view of sample vectors to be evaluated. The matrix has dimensions (dim, lamb) with sample vectors as columns.

```rust
pub fn x(&self) -> DMatrixView<'_, f64>
```

--------------------------------

### Optimizing Rosenbrock Function with CrfmnesOptimizer

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Example demonstrating how to use CrfmnesOptimizer to optimize the 40-dimensional Rosenbrock test function. It initializes the optimizer, iteratively asks for trial points, evaluates them using the Rosenbrock function, and updates the optimizer with the results.

```rust
use rand::{thread_rng, Rng, SeedableRng};
use rand_xoshiro::Xoroshiro128PlusPlus;
use nalgebra::DVector;
use crfmnes::{rec_lamb, CrfmnesOptimizer, test_functions::rosenbrock};

let mut rng = Xoroshiro128PlusPlus::seed_from_u64(thread_rng().gen());
let dim = 40;
let start_m = DVector::zeros(dim);
let start_sigma = 10.0;
let mut opt = CrfmnesOptimizer::new(start_m.clone(), start_sigma, rec_lamb(dim), &mut rng);

let mut best = f64::INFINITY;
let mut best_x = start_m;

for i in 0..10000 {
    let mut trial = opt.ask(&mut rng);

    let mut evs = Vec::new();
    for (i, x) in trial.x().column_iter().enumerate() {
        let eval = rosenbrock(x.as_slice(), 1.0, 100.0);
        evs.push(eval);
        if eval < best {
            best = eval;
            best_x = x.into_owned();
        }
    }

    trial.tell(evs).unwrap();

    if best < 0.001 {
        break;
    }
}
println!("best: {} best_x: {}", best, best_x);
panic!();

```

--------------------------------

### Get Output Type - Same::Output

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Associated type representing the output type, which should always be Self for the Same trait.

```rust
type Output = T
```

--------------------------------

### Get Type ID - Any::type_id

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Gets the TypeId of the current type. This is part of the Any trait implementation.

```rust
fn type_id(&self) -> TypeId
```

--------------------------------

### CrfmnesOptimizer::ask Method

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Requests a new set of sample points from the optimizer for evaluation. Returns a `Trial` object containing the points.

```rust
pub fn ask<'a, R: Rng>(&'a mut self, rand: &mut R) -> Trial<'a>
```

--------------------------------

### CrfmnesOptimizer::with_v_D Initialization

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Creates a new CrfmnesOptimizer with explicit control over the principal vector 'v' and diagonal scaling matrix 'D'. This allows for more fine-grained control over the sampling distribution.

```rust
pub fn with_v_D( m: DVector<f64>, sigma: f64, v: DVector<f64>, D: DVector<f64>, lamb: usize, ) -> Self
```

--------------------------------

### Trial Struct

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/all.html

Represents a trial in an optimization process.

```APIDOC
## Struct: Trial

### Description
Represents a single trial or iteration within an optimization process.

### Fields
(No specific fields documented in the source)
```

--------------------------------

### CrfmnesOptimizer::new

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Creates a new CrfmnesOptimizer with provided parameters and state. Default initializations are used for D and v.

```APIDOC
## CrfmnesOptimizer::new

### Description
Create a new optimiser with the provided parameters and state. Default initialisation is used for `D` and `v`. `D` is set to the identity matrix and `v` is set to a small random vector. See `with_v_D` for more details.

### Parameters
* **m** (DVector<f64>) - The initial mean position vector.
* **sigma** (f64) - The initial size, standard deviation, of the sampling distribution.
* **lamb** (usize) - The number of sample vectors generated for each trial.
* **rand** (&mut R) - A mutable reference to a random number generator.
```

--------------------------------

### Trial::x

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Retrieves a matrix of sample vectors to be evaluated.

```APIDOC
## Trial::x

```rust
pub fn x(&self) -> DMatrixView<'_, f64>
```

### Description
A matrix of shape (dim, lamb) containing sample vectors to be evaluated as columns.
```

--------------------------------

### Trial::tell

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Reports the results of evaluating sample points and updates the optimiser's internal state.

```APIDOC
## Trial::tell

```rust
pub fn tell(&mut self, evs: Vec<f64>) -> Result<(), TrialError>
```

### Description
Tell the optimiser the results of evaluating the sample points, updating its internal state based on the results of this trial.
The user provided evaluations in `evs` should be in the same order as the columns of `x`.
If an error is returned, the state of the parent optimiser is not updated, and a new trial can be attempted.

### Panics
If `evs.len() != x.len()` this method will panic.
```

--------------------------------

### CrfmnesOptimizer::ask

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Asks the optimizer to supply a new set of sample points for evaluation.

```APIDOC
## CrfmnesOptimizer::ask

### Description
Ask the optimiser to supply a new set of sample points to be evaluated.

### Parameters
* **rand** (&mut R) - A mutable reference to a random number generator.

### Returns
* **Trial<'a>** - A `Trial` object containing the sample points.
```

--------------------------------

### Ownership and Cloning

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Methods for managing data ownership and cloning.

```APIDOC
## type Owned = T

### Description
The resulting type after obtaining ownership.

### Type Alias
`Owned`
```

```APIDOC
## fn to_owned(&self) -> T

### Description
Creates owned data from borrowed data, usually by cloning.

### Method
`to_owned`
```

```APIDOC
## fn clone_into(&self, target: &mut T)

### Description
Uses borrowed data to replace owned data, usually by cloning.

### Method
`clone_into`
```

--------------------------------

### CrfmnesOptimizer::lamb Method

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the number of samples generated per `Trial`.

```rust
pub fn lamb(&self) -> usize
```

--------------------------------

### Report Trial Evaluations - Trial::tell

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Reports the results of evaluating sample points to the optimizer. The evaluations should be in the same order as the columns of `x`. Panics if the number of evaluations does not match the number of sample vectors.

```rust
pub fn tell(&mut self, evs: Vec<f64>) -> Result<(), TrialError>
```

--------------------------------

### Test Functions

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/all.html

A collection of test functions for various optimization algorithms.

```APIDOC
## Module: test_functions

### Description
Provides a suite of test functions, likely used for benchmarking or validating optimization algorithms.

### Functions
- ackley
- ackley_param
- beale
- booth
- rastrigin
- rastrigin_a
- rosenbrock
- rosenbrock_2d
- rosenbrock_2d_derivative
- rosenbrock_2d_hessian
- schaffer_n2
- schaffer_n4
- sphere
- sphere_derivative

(No specific details on parameters or return types for these functions are provided in the source.)
```

--------------------------------

### CrfmnesOptimizer CloneToUninit Trait Implementation

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

A nightly-only experimental API for performing copy-assignment to uninitialized memory.

```rust
unsafe fn clone_to_uninit(&self, dest: *mut u8)
```

--------------------------------

### CrfmnesOptimizer::new Initialization

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Creates a new CrfmnesOptimizer with provided mean vector, initial standard deviation, lambda value, and a random number generator. Default initializations are used for internal state variables 'D' and 'v'.

```rust
pub fn new<R: Rng>( m: DVector<f64>, sigma: f64, lamb: usize, rand: &mut R, ) -> Self
```

--------------------------------

### Trial Struct

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

The Trial struct holds the next set of trial values to be evaluated by the user.

```APIDOC
## Struct Trial

```rust
pub struct Trial<'a>
```

### Description
The next set of trial values to be evaluated by the user.
```

--------------------------------

### Create from Subset - SupersetOf::from_subset

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Converts the subset element to the equivalent element of its superset.

```rust
fn from_subset(element: &SS) -> SP
```

--------------------------------

### CrfmnesOptimizer::with_v_D

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Creates a new CrfmnesOptimizer with specified parameters and state, allowing fine-grained control over the sampling distribution.

```APIDOC
## CrfmnesOptimizer::with_v_D

### Description
Create a new optimiser with the provided parameters and state. The parameters in order of application when generating samples from a standard normal distribution:
  * `lamb` determines the number of sample vectors generated for each trial. If an odd number is provided, the next even number is used. For smooth, uni-modal problems use the value provided by `rec_lamb`.
  * `v` is a principal vector which stretches the sampling distribution in an arbitrary direction.
  * `D` is a diagonal matrix (stored as a vector) which scales the distribution along each axis of the problem.
  * `sigma` is the initial size, standard deviation, of the sampling distribution.
  * `m` is the initial mean position vector of the sampling distribution.

### Parameters
* **m** (DVector<f64>) - The initial mean position vector.
* **sigma** (f64) - The initial size, standard deviation, of the sampling distribution.
* **v** (DVector<f64>) - A principal vector that stretches the sampling distribution.
* **D** (DVector<f64>) - A diagonal matrix (stored as a vector) that scales the distribution along each axis.
* **lamb** (usize) - The number of sample vectors generated for each trial.

### Panics
* If `m.is_empty()`
* If `lamb < 4`
* If `sigma <= 0.0`
* If `m.len() != v.len()`
* If `m.len() != D.len()`
```

--------------------------------

### CrfmnesOptimizer::dim

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the dimensionality of the problem space.

```APIDOC
## CrfmnesOptimizer::dim

### Description
The dimensions of the problem space.

### Returns
* **usize** - The dimensionality of the problem space.
```

--------------------------------

### Type Conversion

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Methods for attempting type conversions.

```APIDOC
## type Error = Infallible

### Description
The type returned in the event of a conversion error.

### Type Alias
`Error`
```

```APIDOC
## fn try_from(value: U) -> Result<T, <T as TryFrom<U>>::Error>

### Description
Performs the conversion.

### Method
`try_from`
```

```APIDOC
## type Error = <U as TryFrom<T>>::Error

### Description
The type returned in the event of a conversion error.

### Type Alias
`Error`
```

```APIDOC
## fn try_into(self) -> Result<U, <U as TryFrom<T>>::Error>

### Description
Performs the conversion.

### Method
`try_into`
```

--------------------------------

### Subset Operations

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Methods for interacting with subsets and supersets.

```APIDOC
## fn to_subset(&self) -> Option<SS>

### Description
The inverse inclusion map: attempts to construct `self` from the equivalent element of its superset.

### Method
`to_subset`
```

```APIDOC
## fn is_in_subset(&self) -> bool

### Description
Checks if `self` is actually part of its subset `T` (and can be converted to it).

### Method
`is_in_subset`
```

```APIDOC
## fn to_subset_unchecked(&self) -> SS

### Description
Use with care! Same as `self.to_subset` but without any property checks. Always succeeds.

### Method
`to_subset_unchecked`
```

```APIDOC
## fn from_subset(element: &SS) -> SP

### Description
The inclusion map: converts `self` to the equivalent element of its superset.

### Method
`from_subset`
```

--------------------------------

### CrfmnesOptimizer::lamb

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the number of samples used per trial.

```APIDOC
## CrfmnesOptimizer::lamb

### Description
The number of samples per `Trial`.

### Returns
* **usize** - The number of samples per trial.
```

--------------------------------

### Create From Value - From::from

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Creates a value from another value. This is a blanket implementation of the From trait.

```rust
fn from(t: T) -> T
```

--------------------------------

### Format Trial for Debugging

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Implements the Debug trait for the Trial struct, allowing it to be formatted for debugging purposes.

```rust
fn fmt(&self, f: &mut Formatter<'_>) -> Result
```

--------------------------------

### Convert to Subset - SupersetOf::to_subset

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Attempts to construct the subset from the superset. Returns None if the conversion is not possible.

```rust
fn to_subset(&self) -> Option<SS>
```

--------------------------------

### CrfmnesOptimizer::dim Method

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the dimensionality of the problem space the optimizer is working in.

```rust
pub fn dim(&self) -> usize
```

--------------------------------

### CrfmnesOptimizer::update_count Method

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the number of successful updates that have been performed using the `Trial::tell` method.

```rust
pub fn update_count(&self) -> usize
```

--------------------------------

### CrfmnesOptimizer CloneFrom Implementation

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Enables efficient copying of state from one `CrfmnesOptimizer` to another.

```rust
fn clone_from(&mut self, source: &Self)
```

--------------------------------

### CrfmnesOptimizer Clone Implementation

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Provides the ability to create a duplicate of a `CrfmnesOptimizer` instance.

```rust
fn clone(&self) -> CrfmnesOptimizer
```

--------------------------------

### schaffer_n2

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.schaffer_n2.html

The Schaffer test function No. 2 is defined as `f(x_1, x_2) = 0.5 + (sin^2(x_1^2 - x_2^2) - 0.5) / (1 + 0.001*(x_1^2 + x_2^2))^2`. The input `param` is a slice of f64 representing the input variables (expected to be `[x_1, x_2]`), and the function returns a single f64 value which is the result of the function evaluation. The domain for `x_i` is `[-100, 100]`, and the global minimum is at `f(0, 0) = 0`.

```APIDOC
## Function schaffer_n2

### Description
Evaluates the Schaffer test function No. 2.

### Signature
```rust
pub fn schaffer_n2(param: &[f64]) -> f64
```

### Parameters
* `param` (*slice of f64*) - Input parameters for the function. Expected to contain two elements `[x_1, x_2]` where `x_i` are in the range `[-100, 100]`.

### Returns
* (*f64*) - The result of the Schaffer N.2 function evaluation.

### Mathematical Definition
`f(x_1, x_2) = 0.5 + (sin^2(x_1^2 - x_2^2) - 0.5) / (1 + 0.001*(x_1^2 + x_2^2))^2`

### Global Minimum
The global minimum is at `f(0, 0) = 0`.
```

--------------------------------

### beale

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.beale.html

The Beale test function is defined as f(x1, x2) = (1.5 - x1 + x1*x2)^2 + (2.25 - x1 + x1*x2^2)^2 + (2.625 - x1 + x1*x2^3)^2. The function expects a slice of f64 as input and returns a single f64 value. The domain for x_i is [-4.5, 4.5], and the global minimum is at f(3, 0.5) = 0.

```APIDOC
## Function beale

### Description
Beale test function.

### Signature
```rust
pub fn beale(param: &[f64]) -> f64
```

### Mathematical Definition
`f(x_1, x_2) = (1.5 - x_1 + x_1 * x_2)^2 + (2.25 - x_1 + x_1 * x_2^2)^2 + (2.625 - x_1 + x1 * x_2^3)^2`

### Domain
`x_i \in [-4.5, 4.5]`

### Global Minimum
At `f(x_1, x_2) = f(3, 0.5) = 0`
```

--------------------------------

### VZip Operation

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Method for VZip operation.

```APIDOC
## fn vzip(self) -> V

### Description
Performs the VZip operation.

### Method
`vzip`
```

--------------------------------

### CrfmnesOptimizer Struct

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/all.html

Represents an optimizer within the crfmnes library.

```APIDOC
## Struct: CrfmnesOptimizer

### Description
Represents an optimizer. This is a core data structure for managing optimization processes.

### Fields
(No specific fields documented in the source)
```

--------------------------------

### Try Convert From - TryFrom::try_from

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Performs a fallible conversion from one type to another. The associated Error type is Infallible.

```rust
fn try_from(value: U) -> Result<T, <T as TryFrom<U>>::Error>
```

--------------------------------

### rastrigin

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rastrigin.html

The Rastrigin function is a non-convex, multimodal function used as a performance test for optimization algorithms. It is known for its large number of local optima.

```APIDOC
## rastrigin

### Description
Rastrigin test function.

### Signature
```rust
pub fn rastrigin(param: &[f64]) -> f64
```

### Parameters
* **param** (`&[f64]`) - A slice of f64 representing the input parameters to the function.
```

--------------------------------

### TrialError Enum

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/all.html

Defines possible errors that can occur during a trial.

```APIDOC
## Enum: TrialError

### Description
Enumerates the different types of errors that can occur during an optimization trial.

### Variants
(No specific variants documented in the source)
```

--------------------------------

### rastrigin_a

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rastrigin_a.html

Calculates the Rastrigin test function with a customizable parameter 'a'.

```APIDOC
## rastrigin_a

### Description

Rastrigin test function. The same as `rastrigin`; however, it allows to set the parameter a.

### Function Signature

```rust
pub fn rastrigin_a(param: &[f64], a: f64) -> f64
```

### Parameters

- **param** (`&[f64]`) - The input array of floating-point numbers.
- **a** (`f64`) - The customizable parameter 'a'.
```

--------------------------------

### CrfmnesOptimizer::update_count

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Returns the number of successful updates performed via the `Trial::tell` method.

```APIDOC
## CrfmnesOptimizer::update_count

### Description
Returns the number of sucessful updates performed via the `Trial::tell` method.

### Returns
* **usize** - The count of successful updates.
```

--------------------------------

### rec_lamb Function

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/all.html

A function named rec_lamb.

```APIDOC
## Function: rec_lamb

### Description
This function is named rec_lamb.

### Signature
(No specific signature documented in the source)
```

--------------------------------

### Try Convert Into - TryInto::try_into

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Performs a fallible conversion into another type. The associated Error type is determined by the target type's TryFrom implementation.

```rust
fn try_into(self) -> Result<U, <U as TryFrom<T>>::Error>
```

--------------------------------

### Define Trial Struct

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Defines the Trial struct, which holds private fields and is used to represent a set of trial values for evaluation.

```rust
pub struct Trial<'a> { /* private fields */ }
```

--------------------------------

### Sphere Function Signature

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.sphere.html

Defines the signature for the Sphere test function. It takes a slice of f64 as input and returns a single f64 value.

```rust
pub fn sphere(param: &[f64]) -> f64
```

--------------------------------

### ackley_param

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.ackley_param.html

Calculates the Ackley test function with customizable parameters a, b, and c.

```APIDOC
## Function ackley_param

### Description
Calculates the Ackley test function. This is a variant of the standard `ackley` function that allows for the explicit setting of parameters a, b, and c.

### Signature
```rust
pub fn ackley_param(param: &[f64], a: f64, b: f64, c: f64) -> f64
```

### Parameters
* **param** (`&[f64]`) - The input parameters for the Ackley function.
* **a** (`f64`) - The parameter 'a' for the Ackley function.
* **b** (`f64`) - The parameter 'b' for the Ackley function.
* **c** (`f64`) - The parameter 'c' for the Ackley function.

### Returns
* (`f64`) - The result of the Ackley test function.
```

--------------------------------

### CrfmnesOptimizer Struct Definition

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.CrfmnesOptimizer.html

Defines the CrfmnesOptimizer struct, which is a high-dimension black-box optimizer with an ask-tell interface. It has private fields.

```rust
pub struct CrfmnesOptimizer { /* private fields */ }
```

--------------------------------

### rec_lamb Function Signature

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/fn.rec_lamb.html

Calculates the recommended lambda value for a given dimension size. Higher values may be suitable for noisy or multi-modal objective functions.

```APIDOC
## rec_lamb

### Description
Recommended lambda for a given dim size for typical problems. Noisy or highly multi-modal objective functions should use higher values, e.g. 4*dim.

### Signature
```rust
pub fn rec_lamb(dim: usize) -> usize
```

### Parameters
#### Path Parameters
- **dim** (usize) - The dimension size for which to calculate the recommended lambda.
```

--------------------------------

### sphere_derivative

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.sphere_derivative.html

Calculates the derivative of the sphere test function. The function is defined as f(x_1, x_2, ..., x_n) = (2 * x_1, 2 * x_2, ... 2 * x_n), where x_i are real numbers and n > 0.

```APIDOC
## Function sphere_derivative

### Description
Calculates the derivative of the sphere test function.

Defined as:
`f(x_1, x_2, ..., x_n) = (2 * x_1, 2 * x_2, ... 2 * x_n)`
where `x_i \in (-\infty, \infty)` and `n > 0`.

### Signature
```rust
pub fn sphere_derivative(param: &[f64]) -> Vec<f64>
```

### Parameters
* `param` (`&[f64]`) - A slice of f64 representing the input vector x.
```

--------------------------------

### rosenbrock_2d_hessian function signature

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rosenbrock_2d_hessian.html

This is the function signature for the rosenbrock_2d_hessian function. It takes a slice of f64 parameters and a scalar b, returning a Vec<f64> representing the Hessian.

```rust
pub fn rosenbrock_2d_hessian(param: &[f64], _a: f64, b: f64) -> Vec<f64>
```

--------------------------------

### Convert Into Value - Into::into

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Converts a value into another type. This is a blanket implementation of the Into trait.

```rust
fn into(self) -> U
```

--------------------------------

### ackley

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.ackley.html

The ackley function computes the Ackley test function value for a given set of parameters. The function is defined mathematically and has a global minimum at zero.

```APIDOC
## ackley

### Description
Computes the Ackley test function value.

### Signature
```rust
pub fn ackley(param: &[f64]) -> f64
```

### Mathematical Definition
`f(x_1, x_2, ..., x_n) = - a * exp( -b \sqrt{\frac{1}{d}\sum_{i=1}^n x_i^2 } ) - exp( \frac{1}{d} cos(c * x_i) ) + a + exp(1)`

where `x_i \in [-32.768, 32.768]` and usually `a = 10`, `b = 0.2` and `c = 2*pi`.

### Global Minimum
The global minimum is at `f(0, 0, ..., 0) = 0`.
```

--------------------------------

### Convert to Subset Unchecked - SupersetOf::to_subset_unchecked

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Converts to the subset without property checks. Use with care as it always succeeds.

```rust
fn to_subset_unchecked(&self) -> SS
```

--------------------------------

### Define Enum TrialError

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/enum.TrialError.html

Defines the possible error types that can be returned by `Trial::tell`. In all cases, the trial is discarded and can be retried.

```rust
pub enum TrialError {
    NoFeasibleSolutions,
    DivByZero,
    DiagonalInverted,
}
```

--------------------------------

### Check if in Subset - SupersetOf::is_in_subset

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Checks if the current value is part of its subset and can be converted.

```rust
fn is_in_subset(&self) -> bool
```

--------------------------------

### rosenbrock_2d

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rosenbrock_2d.html

Calculates the value of the 2D Rosenbrock test function. The function is defined as f(x_1, x_2) = (a - x_1)^2 + b * (x_2 - x_1^2)^2. The global minimum is at f(1, 1) = 0.

```APIDOC
## Function rosenbrock_2d

### Description
Calculates the value of the 2D Rosenbrock test function.
Defined as `f(x_1, x_2) = (a - x_1)^2 + b * (x_2 - x_1^2)^2`
where `x_i \in (-\infty, \infty)`. The parameters a and b usually are: `a = 1` and `b = 100`.
For 2D problems, this function is much faster than `rosenbrock`.
The global minimum is at `f(x_1, x_2) = f(1, 1) = 0`.

### Signature
```rust
pub fn rosenbrock_2d(param: &[f64], a: f64, b: f64) -> f64
```

### Parameters
* `param` (array of f64): The input parameters [x1, x2].
* `a` (f64): The parameter 'a' in the Rosenbrock function definition.
* `b` (f64): The parameter 'b' in the Rosenbrock function definition.

### Returns
* `f64`: The computed value of the Rosenbrock 2D function.
```

--------------------------------

### rosenbrock_2d_derivative

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rosenbrock_2d_derivative.html

Calculates the derivative of the 2D Rosenbrock function. The function takes a slice of f64 as the first parameter, representing the input point, and two f64 values 'a' and 'b' as parameters for the Rosenbrock function. It returns a Vec<f64> representing the gradient.

```APIDOC
## rosenbrock_2d_derivative

### Description
Derivative of 2D Rosenbrock function.

### Signature
```rust
pub fn rosenbrock_2d_derivative(param: &[f64], a: f64, b: f64) -> Vec<f64>
```

### Parameters
* **param** (`&[f64]`) - The input point for the function.
* **a** (`f64`) - Parameter 'a' for the Rosenbrock function.
* **b** (`f64`) - Parameter 'b' for the Rosenbrock function.

### Returns
* `Vec<f64>` - The gradient of the 2D Rosenbrock function at the given point.
```

--------------------------------

### Borrow Value - Borrow::borrow

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Immutably borrows the value. This is part of the Borrow trait implementation.

```rust
fn borrow(&self) -> &T
```

--------------------------------

### sphere

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.sphere.html

Calculates the value of the sphere test function. This function is defined as the sum of the squares of its input parameters.

```APIDOC
## sphere

### Description
Calculates the value of the sphere test function.

### Function Signature
```rust
pub fn sphere(param: &[f64]) -> f64
```

### Parameters
* **param** (slice of f64) - The input parameters for the function. Each element `x_i` can be any real number.

### Returns
* `f64` - The calculated value of the sphere function, which is the sum of the squares of the input parameters.

### Mathematical Definition
`f(x_1, x_2, …, x_n) = \sum_{i=1}^n x_i^2`
where `x_i \in (-\infty, \infty)` and `n > 0`.

### Global Minimum
The global minimum is at `f(0, 0, ..., 0) = 0`.
```

--------------------------------

### Vector Zip - VZip::vzip

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Performs a vector zip operation. This is part of the VZip trait implementation.

```rust
fn vzip(self) -> V
```

--------------------------------

### rosenbrock_2d_hessian

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rosenbrock_2d_hessian.html

Calculates the Hessian of the 2D Rosenbrock function. This function is part of the crfmnes::test_functions module.

```APIDOC
## Function rosenbrock_2d_hessian

### Description
Hessian of 2D Rosenbrock function

### Signature
```rust
pub fn rosenbrock_2d_hessian(param: &[f64], _a: f64, b: f64) -> Vec<f64>
```

### Parameters
* `param` (`&[f64]`) - Input parameters for the function.
* `_a` (`f64`) - Parameter 'a' for the Rosenbrock function (unused in Hessian calculation).
* `b` (`f64`) - Parameter 'b' for the Rosenbrock function.

### Returns
* `Vec<f64>` - A vector representing the Hessian matrix.
```

--------------------------------

### Mutably Borrow Value - BorrowMut::borrow_mut

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/struct.Trial.html

Mutably borrows the value. This is part of the BorrowMut trait implementation.

```rust
fn borrow_mut(&mut self) -> &mut T
```

--------------------------------

### Rosenbrock Function Signature

Source: https://docs.rs/crfmnes/1.0.0/crfmnes/test_functions/fn.rosenbrock.html

The rosenbrock function calculates the value of the multidimensional Rosenbrock test function. It takes a slice of f64 values representing the coordinates and two f64 parameters, a and b, and returns a single f64 value.

```APIDOC
## rosenbrock

### Description

Calculates the value of the multidimensional Rosenbrock test function.

Defined as:
`f(x_1, x_2, ..., x_n) = \sum_{i=1}^{n-1} \left[ (a - x_i)^2 + b * (x_{i+1} - x_i^2)^2 \right]`
where `x_i \in (-\infty, \infty)`. The parameters a and b usually are: `a = 1` and `b = 100`.

The global minimum is at `f(x_1, x_2, ..., x_n) = f(1, 1, ..., 1) = 0`.

### Function Signature

```rust
pub fn rosenbrock(param: &[f64], a: f64, b: f64) -> f64
```

### Parameters

- **param** (*&[f64]*) - Required - A slice of f64 values representing the coordinates (x_1, x_2, ..., x_n).
- **a** (*f64*) - Required - The parameter 'a' in the Rosenbrock function definition.
- **b** (*f64*) - Required - The parameter 'b' in the Rosenbrock function definition.

### Returns

- **f64** - The calculated value of the Rosenbrock function.
```
Tokens:
Raw