1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
extern crate ndarray;
extern crate ndarray_linalg;
use ndarray::*;
use ndarray_rand::RandomExt;
use ndarray_rand::rand_distr::StandardNormal;
use crate::feature_collection::*;
use crate::params::*;
#[derive(Clone)]
pub struct SketchedLinearFeatureCollection {
in_dimensions : usize,
out_dimensions : usize,
alpha : f32,
projection_mat : Array2<f32>
}
impl SketchedLinearFeatureCollection {
pub fn new(in_dimensions : usize, out_dimensions : usize, alpha : f32) -> SketchedLinearFeatureCollection {
let projection_mat = Array::random((out_dimensions, in_dimensions), StandardNormal);
SketchedLinearFeatureCollection {
in_dimensions,
out_dimensions,
alpha,
projection_mat
}
}
}
impl FeatureCollection for SketchedLinearFeatureCollection {
fn get_in_dimensions(&self) -> usize {
self.in_dimensions
}
fn get_dimension(&self) -> usize {
self.out_dimensions + 1
}
fn get_features(&self, in_vec: ArrayView1<f32>) -> Array1<f32> {
let projected = self.projection_mat.dot(&in_vec);
let single_ones = Array::ones((1,));
let result = stack(Axis(0), &[projected.view(), single_ones.view()]).unwrap();
self.alpha * result
}
fn get_jacobian(&self, _in_vec : ArrayView1<f32>) -> Array2<f32> {
let zero_row = Array::zeros((1,self.in_dimensions));
let result = stack(Axis(0), &[self.projection_mat.view(), zero_row.view()]).unwrap();
self.alpha * result
}
}