scirs2_integrate/autodiff/
mod.rs

1//! Enhanced automatic differentiation for numerical integration
2//!
3//! This module provides advanced automatic differentiation capabilities including:
4//! - Forward mode AD for efficient gradient computation
5//! - Reverse mode AD for efficient Jacobian computation
6//! - Sparse Jacobian optimization
7//! - Sensitivity analysis tools
8
9pub mod dual;
10pub mod forward;
11pub mod reverse;
12pub mod sensitivity;
13pub mod sparse;
14
15// Re-export main types and functions
16pub use dual::{Dual, DualVector};
17pub use forward::{
18    forward_gradient, forward_jacobian, ForwardAD, ForwardODEJacobian, VectorizedForwardAD,
19};
20pub use reverse::{
21    reverse_gradient, reverse_jacobian, CheckpointStrategy, ReverseAD, Tape, TapeNode,
22};
23pub use sensitivity::{
24    compute_sensitivities, MorrisScreening, ParameterSensitivity, SensitivityAnalysis,
25    SobolSensitivity, EFAST,
26};
27pub use sparse::{
28    colored_jacobian, compress_jacobian, detect_sparsity, detect_sparsity_adaptive, BlockPattern,
29    CSCJacobian, CSRJacobian, ColGrouping, HybridJacobian, SparseJacobian, SparseJacobianUpdater,
30    SparsePattern,
31};
scirs2_integrate/autodiff/mod.rs

scirs2_integrate/autodiff/
mod.rs
