work/smartcore
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: documents matrix methods

Browse Source
This commit is contained in:
Volodymyr Orlov
2020-09-06 18:27:11 -07:00
parent 1e3ed4c924
commit bbe810d164
25 changed files with 587 additions and 245 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/optimization/first_order/gradient_descent.rs
+3 -3
View File
@@ -58,10 +58,10 @@ impl<T: RealNumber> FirstOrderOptimizer<T> for GradientDescent<T> {
let mut dg = gvec.clone();
dx.mul_scalar_mut(alpha);
df(&mut dg, &dx.add_mut(&x)); //df(x) = df(x .+ gvec .* alpha)
gvec.vector_dot(&dg)
gvec.dot(&dg)
};
let df0 = step.vector_dot(&gvec);
let df0 = step.dot(&gvec);
let ls_r = ls.search(&f_alpha, &df_alpha, alpha, fx, df0);
alpha = ls_r.alpha;
@@ -90,7 +90,7 @@ mod tests {
#[test]
fn gradient_descent() {
let x0 = DenseMatrix::vector_from_array(&[-1., 1.]);
let x0 = DenseMatrix::row_vector_from_array(&[-1., 1.]);
let f = |x: &DenseMatrix<f64>| {
(1.0 - x.get(0, 0)).powf(2.) + 100.0 * (x.get(0, 1) - x.get(0, 0).powf(2.)).powf(2.)
};
src/optimization/first_order/lbfgs.rs
+7 -7
View File
@@ -46,7 +46,7 @@ impl<T: RealNumber> LBFGS<T> {
let i = index.rem_euclid(self.m);
let dgi = &state.dg_history[i];
let dxi = &state.dx_history[i];
state.twoloop_alpha[i] = state.rho[i] * dxi.vector_dot(&state.twoloop_q);
state.twoloop_alpha[i] = state.rho[i] * dxi.dot(&state.twoloop_q);
state
.twoloop_q
.sub_mut(&dgi.mul_scalar(state.twoloop_alpha[i]));
@@ -56,7 +56,7 @@ impl<T: RealNumber> LBFGS<T> {
let i = (upper - 1).rem_euclid(self.m);
let dxi = &state.dx_history[i];
let dgi = &state.dg_history[i];
let scaling = dxi.vector_dot(dgi) / dgi.abs().pow_mut(T::two()).sum();
let scaling = dxi.dot(dgi) / dgi.abs().pow_mut(T::two()).sum();
state.s.copy_from(&state.twoloop_q.mul_scalar(scaling));
} else {
state.s.copy_from(&state.twoloop_q);
@@ -66,7 +66,7 @@ impl<T: RealNumber> LBFGS<T> {
let i = index.rem_euclid(self.m);
let dgi = &state.dg_history[i];
let dxi = &state.dx_history[i];
let beta = state.rho[i] * dgi.vector_dot(&state.s);
let beta = state.rho[i] * dgi.dot(&state.s);
state
.s
.add_mut(&dxi.mul_scalar(state.twoloop_alpha[i] - beta));
@@ -111,7 +111,7 @@ impl<T: RealNumber> LBFGS<T> {
state.x_f_prev = f(&state.x);
state.x_prev.copy_from(&state.x);
let df0 = state.x_df.vector_dot(&state.s);
let df0 = state.x_df.dot(&state.s);
let f_alpha = |alpha: T| -> T {
let mut dx = state.s.clone();
@@ -124,7 +124,7 @@ impl<T: RealNumber> LBFGS<T> {
let mut dg = state.x_df.clone();
dx.mul_scalar_mut(alpha);
df(&mut dg, &dx.add_mut(&state.x)); //df(x) = df(x .+ gvec .* alpha)
state.x_df.vector_dot(&dg)
state.x_df.dot(&dg)
};
let ls_r = ls.search(&f_alpha, &df_alpha, T::one(), state.x_f_prev, df0);
@@ -164,7 +164,7 @@ impl<T: RealNumber> LBFGS<T> {
fn update_hessian<'a, X: Matrix<T>>(&self, _: &'a DF<X>, state: &mut LBFGSState<T, X>) {
state.dg = state.x_df.sub(&state.x_df_prev);
let rho_iteration = T::one() / state.dx.vector_dot(&state.dg);
let rho_iteration = T::one() / state.dx.dot(&state.dg);
if !rho_iteration.is_infinite() {
let idx = state.iteration.rem_euclid(self.m);
state.dx_history[idx].copy_from(&state.dx);
@@ -240,7 +240,7 @@ mod tests {
#[test]
fn lbfgs() {
let x0 = DenseMatrix::vector_from_array(&[0., 0.]);
let x0 = DenseMatrix::row_vector_from_array(&[0., 0.]);
let f = |x: &DenseMatrix<f64>| {
(1.0 - x.get(0, 0)).powf(2.) + 100.0 * (x.get(0, 1) - x.get(0, 0).powf(2.)).powf(2.)
};