2018-10-21 04:26:44 +08:00
|
|
|
use alga::general::{ClosedAdd, ClosedMul};
|
|
|
|
use num::{One, Zero};
|
2018-10-21 13:42:32 +08:00
|
|
|
use std::iter;
|
2018-10-21 04:26:44 +08:00
|
|
|
use std::marker::PhantomData;
|
|
|
|
use std::ops::{Add, Mul, Range};
|
2018-10-21 13:42:32 +08:00
|
|
|
use std::slice;
|
2018-10-21 04:26:44 +08:00
|
|
|
|
|
|
|
use allocator::Allocator;
|
2018-10-24 00:18:05 +08:00
|
|
|
use constraint::{AreMultipliable, DimEq, SameNumberOfRows, ShapeConstraint};
|
2018-10-21 04:26:44 +08:00
|
|
|
use storage::{Storage, StorageMut};
|
2018-10-24 00:18:05 +08:00
|
|
|
use {DefaultAllocator, Dim, Matrix, MatrixMN, Real, Scalar, Vector, VectorN, U1};
|
2018-10-21 04:26:44 +08:00
|
|
|
|
2018-10-21 13:42:32 +08:00
|
|
|
// FIXME: this structure exists for now only because impl trait
|
|
|
|
// cannot be used for trait method return types.
|
|
|
|
pub trait CsStorageIter<'a, N, R, C = U1> {
|
|
|
|
type ColumnEntries: Iterator<Item = (usize, N)>;
|
2018-10-30 14:46:34 +08:00
|
|
|
type ColumnRowIndices: Iterator<Item = usize>;
|
2018-10-21 13:42:32 +08:00
|
|
|
|
2018-10-30 14:46:34 +08:00
|
|
|
fn column_row_indices(&'a self, j: usize) -> Self::ColumnRowIndices;
|
2018-10-21 13:42:32 +08:00
|
|
|
fn column_entries(&'a self, j: usize) -> Self::ColumnEntries;
|
|
|
|
}
|
|
|
|
|
|
|
|
pub trait CsStorage<N, R, C = U1>: for<'a> CsStorageIter<'a, N, R, C> {
|
2018-10-21 04:26:44 +08:00
|
|
|
fn shape(&self) -> (R, C);
|
|
|
|
unsafe fn row_index_unchecked(&self, i: usize) -> usize;
|
|
|
|
unsafe fn get_value_unchecked(&self, i: usize) -> &N;
|
|
|
|
fn get_value(&self, i: usize) -> &N;
|
|
|
|
fn row_index(&self, i: usize) -> usize;
|
2018-10-24 00:18:05 +08:00
|
|
|
fn column_range(&self, i: usize) -> Range<usize>;
|
|
|
|
fn len(&self) -> usize;
|
2018-10-21 04:26:44 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
pub trait CsStorageMut<N, R, C = U1>: CsStorage<N, R, C> {
|
|
|
|
/*
|
|
|
|
/// Sets the length of this column without initializing its values and row indices.
|
|
|
|
///
|
|
|
|
/// If the given length is larger than the current one, uninitialized entries are
|
|
|
|
/// added at the end of the column `i`. This will effectively shift all the matrix entries
|
2018-10-21 13:42:32 +08:00
|
|
|
/// of the columns at indices `j` with `j > i`.
|
|
|
|
fn set_column_len(&mut self, i: usize, len: usize);
|
2018-10-21 04:26:44 +08:00
|
|
|
*/
|
|
|
|
}
|
|
|
|
|
|
|
|
#[derive(Clone, Debug)]
|
|
|
|
pub struct CsVecStorage<N: Scalar, R: Dim, C: Dim>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, C>,
|
|
|
|
{
|
2018-10-30 14:46:34 +08:00
|
|
|
pub(crate) shape: (R, C),
|
|
|
|
pub(crate) p: VectorN<usize, C>,
|
|
|
|
pub(crate) i: Vec<usize>,
|
|
|
|
pub(crate) vals: Vec<N>,
|
2018-10-21 04:26:44 +08:00
|
|
|
}
|
|
|
|
|
2018-10-31 00:29:32 +08:00
|
|
|
impl<N: Scalar, R: Dim, C: Dim> CsVecStorage<N, R, C>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, C>,
|
|
|
|
{
|
|
|
|
pub fn values(&self) -> &[N] {
|
|
|
|
&self.vals
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-10-24 00:18:05 +08:00
|
|
|
impl<N: Scalar, R: Dim, C: Dim> CsVecStorage<N, R, C> where DefaultAllocator: Allocator<usize, C> {}
|
2018-10-21 13:42:32 +08:00
|
|
|
|
|
|
|
impl<'a, N: Scalar, R: Dim, C: Dim> CsStorageIter<'a, N, R, C> for CsVecStorage<N, R, C>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, C>,
|
|
|
|
{
|
|
|
|
type ColumnEntries =
|
|
|
|
iter::Zip<iter::Cloned<slice::Iter<'a, usize>>, iter::Cloned<slice::Iter<'a, N>>>;
|
2018-10-30 14:46:34 +08:00
|
|
|
type ColumnRowIndices = iter::Cloned<slice::Iter<'a, usize>>;
|
2018-10-21 13:42:32 +08:00
|
|
|
|
|
|
|
#[inline]
|
|
|
|
fn column_entries(&'a self, j: usize) -> Self::ColumnEntries {
|
|
|
|
let rng = self.column_range(j);
|
|
|
|
self.i[rng.clone()]
|
|
|
|
.iter()
|
|
|
|
.cloned()
|
|
|
|
.zip(self.vals[rng].iter().cloned())
|
|
|
|
}
|
2018-10-30 14:46:34 +08:00
|
|
|
|
|
|
|
#[inline]
|
|
|
|
fn column_row_indices(&'a self, j: usize) -> Self::ColumnRowIndices {
|
|
|
|
let rng = self.column_range(j);
|
|
|
|
self.i[rng.clone()].iter().cloned()
|
|
|
|
}
|
2018-10-21 13:42:32 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
impl<N: Scalar, R: Dim, C: Dim> CsStorage<N, R, C> for CsVecStorage<N, R, C>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, C>,
|
|
|
|
{
|
|
|
|
#[inline]
|
|
|
|
fn shape(&self) -> (R, C) {
|
|
|
|
self.shape
|
|
|
|
}
|
|
|
|
|
|
|
|
#[inline]
|
2018-10-24 00:18:05 +08:00
|
|
|
fn len(&self) -> usize {
|
2018-10-21 13:42:32 +08:00
|
|
|
self.vals.len()
|
|
|
|
}
|
2018-10-21 04:26:44 +08:00
|
|
|
|
|
|
|
#[inline]
|
|
|
|
fn row_index(&self, i: usize) -> usize {
|
|
|
|
self.i[i]
|
|
|
|
}
|
|
|
|
|
|
|
|
#[inline]
|
|
|
|
unsafe fn row_index_unchecked(&self, i: usize) -> usize {
|
|
|
|
*self.i.get_unchecked(i)
|
|
|
|
}
|
|
|
|
|
|
|
|
#[inline]
|
|
|
|
unsafe fn get_value_unchecked(&self, i: usize) -> &N {
|
|
|
|
self.vals.get_unchecked(i)
|
|
|
|
}
|
|
|
|
|
|
|
|
#[inline]
|
|
|
|
fn get_value(&self, i: usize) -> &N {
|
|
|
|
&self.vals[i]
|
|
|
|
}
|
2018-10-24 00:18:05 +08:00
|
|
|
|
|
|
|
#[inline]
|
|
|
|
fn column_range(&self, j: usize) -> Range<usize> {
|
|
|
|
let end = if j + 1 == self.p.len() {
|
|
|
|
self.len()
|
|
|
|
} else {
|
|
|
|
self.p[j + 1]
|
|
|
|
};
|
|
|
|
|
|
|
|
self.p[j]..end
|
|
|
|
}
|
2018-10-21 04:26:44 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
pub struct CsSliceStorage<'a, N: Scalar, R: Dim, C: DimAdd<U1>> {
|
|
|
|
shape: (R, C),
|
|
|
|
p: VectorSlice<usize, DimSum<C, U1>>,
|
|
|
|
i: VectorSlice<usize, Dynamic>,
|
|
|
|
vals: VectorSlice<N, Dynamic>,
|
|
|
|
}*/
|
|
|
|
|
|
|
|
/// A compressed sparse column matrix.
|
|
|
|
#[derive(Clone, Debug)]
|
|
|
|
pub struct CsMatrix<N: Scalar, R: Dim, C: Dim, S: CsStorage<N, R, C> = CsVecStorage<N, R, C>> {
|
|
|
|
pub data: S,
|
|
|
|
_phantoms: PhantomData<(N, R, C)>,
|
|
|
|
}
|
|
|
|
|
2018-10-21 04:27:18 +08:00
|
|
|
pub type CsVector<N, R, S = CsVecStorage<N, R, U1>> = CsMatrix<N, R, U1, S>;
|
2018-10-21 04:26:44 +08:00
|
|
|
|
|
|
|
impl<N: Scalar, R: Dim, C: Dim> CsMatrix<N, R, C>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, C>,
|
|
|
|
{
|
|
|
|
pub fn new_uninitialized_generic(nrows: R, ncols: C, nvals: usize) -> Self {
|
|
|
|
let mut i = Vec::with_capacity(nvals);
|
|
|
|
unsafe {
|
|
|
|
i.set_len(nvals);
|
|
|
|
}
|
|
|
|
i.shrink_to_fit();
|
|
|
|
|
|
|
|
let mut vals = Vec::with_capacity(nvals);
|
|
|
|
unsafe {
|
|
|
|
vals.set_len(nvals);
|
|
|
|
}
|
|
|
|
vals.shrink_to_fit();
|
|
|
|
|
|
|
|
CsMatrix {
|
|
|
|
data: CsVecStorage {
|
|
|
|
shape: (nrows, ncols),
|
2018-10-24 00:18:05 +08:00
|
|
|
p: VectorN::zeros_generic(ncols, U1),
|
2018-10-21 04:26:44 +08:00
|
|
|
i,
|
|
|
|
vals,
|
|
|
|
},
|
|
|
|
_phantoms: PhantomData,
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
fn cumsum<D: Dim>(a: &mut VectorN<usize, D>, b: &mut VectorN<usize, D>) -> usize
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, D>,
|
|
|
|
{
|
|
|
|
assert!(a.len() == b.len());
|
|
|
|
let mut sum = 0;
|
|
|
|
|
|
|
|
for i in 0..a.len() {
|
|
|
|
b[i] = sum;
|
|
|
|
sum += a[i];
|
|
|
|
a[i] = b[i];
|
|
|
|
}
|
|
|
|
|
|
|
|
sum
|
|
|
|
}
|
|
|
|
|
|
|
|
impl<N: Scalar, R: Dim, C: Dim, S: CsStorage<N, R, C>> CsMatrix<N, R, C, S> {
|
2018-10-31 00:29:32 +08:00
|
|
|
pub fn from_data(data: S) -> Self {
|
|
|
|
CsMatrix {
|
|
|
|
data,
|
|
|
|
_phantoms: PhantomData,
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-10-24 00:18:05 +08:00
|
|
|
pub fn len(&self) -> usize {
|
|
|
|
self.data.len()
|
2018-10-21 04:26:44 +08:00
|
|
|
}
|
|
|
|
|
2018-10-31 00:29:32 +08:00
|
|
|
pub fn nrows(&self) -> usize {
|
|
|
|
self.data.shape().0.value()
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn ncols(&self) -> usize {
|
|
|
|
self.data.shape().1.value()
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn shape(&self) -> (usize, usize) {
|
|
|
|
let (nrows, ncols) = self.data.shape();
|
|
|
|
(nrows.value(), ncols.value())
|
|
|
|
}
|
|
|
|
|
|
|
|
pub fn is_square(&self) -> bool {
|
|
|
|
let (nrows, ncols) = self.data.shape();
|
|
|
|
nrows.value() == ncols.value()
|
|
|
|
}
|
|
|
|
|
2018-10-21 04:26:44 +08:00
|
|
|
pub fn transpose(&self) -> CsMatrix<N, C, R>
|
|
|
|
where
|
|
|
|
DefaultAllocator: Allocator<usize, R>,
|
|
|
|
{
|
|
|
|
let (nrows, ncols) = self.data.shape();
|
|
|
|
|
2018-10-24 00:18:05 +08:00
|
|
|
let nvals = self.len();
|
2018-10-21 04:26:44 +08:00
|
|
|
let mut res = CsMatrix::new_uninitialized_generic(ncols, nrows, nvals);
|
|
|
|
let mut workspace = Vector::zeros_generic(nrows, U1);
|
|
|
|
|
|
|
|
// Compute p.
|
|
|
|
for i in 0..nvals {
|
|
|
|
let row_id = self.data.row_index(i);
|
|
|
|
workspace[row_id] += 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
let _ = cumsum(&mut workspace, &mut res.data.p);
|
|
|
|
|
|
|
|
// Fill the result.
|
|
|
|
for j in 0..ncols.value() {
|
2018-10-21 13:42:32 +08:00
|
|
|
for (row_id, value) in self.data.column_entries(j) {
|
2018-10-21 04:26:44 +08:00
|
|
|
let shift = workspace[row_id];
|
|
|
|
|
2018-10-21 13:42:32 +08:00
|
|
|
res.data.vals[shift] = value;
|
2018-10-21 04:26:44 +08:00
|
|
|
res.data.i[shift] = j;
|
|
|
|
workspace[row_id] += 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
res
|
|
|
|
}
|
|
|
|
}
|