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 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561
//! The `neuronika` crate provides auto-differentiation and dynamic neural networks.
//!
//! Neuronika is a machine learning framework written in pure Rust, built with a focus on ease of
//! use, fast experimentation and performance.
//!
//! # Highlights
//!
//! * Define by run computational graphs.
//! * Reverse-mode automatic differentiation.
//! * Dynamic neural networks.
//!
//! # Variables
//!
//! The main building blocks of neuronika are *variables* and *differentiable variables*.
//! This means that when using this crate you will be handling and manipulating instances of [`Var`]
//! and [`VarDiff`].
//!
//! Variables are lean and powerful abstractions over the computational graph's nodes. Neuronika
//! empowers you with the ability of imperatively building and differentiating such graphs with
//! minimal amount of code and effort.
//!
//! Both differentiable and non-differentiable variables can be understood as *tensors*. You
//! can perform all the basic arithmetic operations on them, such as: `+`, `-`, `*` and `/`.
//! Refer to [`Var`] and [`VarDiff`] for a complete list of the available operations.
//!
//! It is important to note that cloning variables is extremely memory efficient as only a shallow
//! copy is returned. Cloning a variable is thus the way to go if you need to use it several times.
//!
//! The provided API is linear in thought and minimal as it is carefully tailored around you, the
//! user.
//!
//! ### Quickstart
//!
//! If you’re familiar with Pytorch or Numpy, you will easily follow these example. If not, brace
//! yourself and follow along.
//!
//! First thing first, you should import neuronika.
//!
//! ```
//! use neuronika;
//! ```
//!
//! Neuronika's variables are designed to work with the [`f32`] data type, although this may change in
//! the future, and can be initialized in many ways. In the following, we will show some of
//! the possible alternatives:
//!
//! **With random or constant values**:
//!
//! Here `shape` determines the dimensionality of the output variable.
//! ```
//! let shape = [3, 4];
//!
//! let rand_variable = neuronika::rand(shape);
//! let ones_variable = neuronika::ones(shape);
//! let constant_variable = neuronika::full(shape, 7.);
//!
//! print!("Full variable:\n{}", constant_variable);
//! ```
//!
//! Out:
//!
//! ```text
//! [[7, 7, 7, 7],
//! [7, 7, 7, 7],
//! [7, 7, 7, 7]]
//! ```
//!
//! **From a ndarray array**
//!
//! ```
//! use ndarray::array;
//!
//! let array = array![1., 2.];
//! let x_ndarray = neuronika::from_ndarray(array);
//!
//! print!("From ndarray:\n{}", x_ndarray);
//! ```
//! Out:
//!
//! ```text
//! [1, 2]
//! ```
//!
//! Accessing the underlying data is possible by using [`.data()`](crate::Data):
//!
//! ```
//! let dim = (2, 2);
//!
//! let x = neuronika::rand(dim);
//!
//! assert_eq!(x.data().dim(), dim);
//! ```
//!
//! ## Leaf Variables
//!
//! You can create leaf variables by using one of the many provided functions, such as [`zeros()`],
//! [`ones()`], [`full()`] and [`rand()`]. Refer to the [complete list](#functions) for additional
//! information.
//!
//! Leaf variables are so called because they form the *leaves* of the computational graph, as are
//! not the result of any computation.
//!
//! Every leaf variable is by default created as non-differentiable, to promote it to a
//! *differentiable* leaf, i. e. a variable for which you can compute the gradient, you can use
//! [`.requires_grad()`](Var::requires_grad()).
//!
//! Differentiable leaf variables are leaves that have been promoted. You will encounter them
//! very often in your journey through neuronika as they are the the main components of the
//! neural networks' building blocks. To learn more in detail about those check the
//! [`nn`](module@nn) module.
//!
//! Differentiable leaves hold a gradient, you can access it with [`.grad()`](VarDiff::grad()).
//!
//! ## Differentiability Arithmetic
//!
//! As stated before, you can manipulate variables by performing operations on them; the results of
//! those computations will also be variables, although not leaf ones.
//!
//! The result of an operation between two differentiable variables will also be a differentiable
//! variable and the converse holds for non-differentiable variables. However, things behave
//! slightly differently when an operation is performed between a non-differentiable variable and a
//! differentiable one, as the resulting variable will be differentiable.
//!
//! You can think of differentiability as a *sticky* property. The table that follows is a summary
//! of how differentiability is broadcasted through variables.
//!
//! **Operands** | Var | VarDiff
//! --------------|---------|---------
//! **Var** | Var | VarDiff
//! **VarDiff** | VarDiff | VarDiff
//!
//!
//! ## Differentiable Ancestors
//!
//! The differentiable ancestors of a variable are the differentiable leaves of the graph involved
//! in its computation. Obviously, only [`VarDiff`] can have a set of ancestors.
//!
//! You can gain access, via mutable views, to all the ancestors of a variable by iterating through
//! the vector of [`Param`] returned by [`.parameters()`](VarDiff::parameters()).
//! To gain more insights about the role that such components fulfil in neuronika feel free to check
//! the [`optim`] module.
//!
//! # Computational Graph
//!
//! A computational graph is implicitly created as you write your program. You can differentiate it
//! with respect to some of the differentiable leaves, thus populating their gradients, by using
//! [`.backward()`](VarDiff::backward()).
//!
//! It is important to note that the computational graph is *lazily* evaluated, this means that
//! neuronika decouples the construction of the graph from the actual computation of the nodes'
//! values. You must use `.forward()` in order to obtain the actual result of the computation.
//!
//!```
//! # #[cfg(feature = "blas")]
//! # extern crate blas_src;
//!use neuronika;
//!
//!let x = neuronika::rand(5); //----+
//!let q = neuronika::rand((5, 5)); // |- Those lines build the graph.
//! // |
//!let y = x.clone().vm(q).vv(x); //----+
//! //
//!y.forward(); // After .forward() is called y contains the result.
//!```
//!
//! ## Freeing and keeping the graph
//!
//! By default, computational graphs will persist in the program's memory. If you want or need to be
//! more conservative about that you can wrap any arbitrary subset of the computations in an inner
//! scope. This allows for the corresponding portion of the graph to be freed when the end of
//! the scope is reached by the execution of your program.
//!
//!```
//! # #[cfg(feature = "blas")]
//! # extern crate blas_src;
//!use neuronika;
//!
//!let w = neuronika::rand((3, 3)).requires_grad(); // -----------------+
//!let b = neuronika::rand(3).requires_grad(); // |
//!let x = neuronika::rand((10, 3)); // -----------------+- Leaves are created
//! //
//!{ // ---+
//! let h = x.mm(w.t()) + b; // | w's and b's
//! h.forward(); // | grads are
//! h.backward(1.0); // | accumulated
//!} // ---+ |- Graph is freed and
//! // -----------------+ only leaves remain
//!```
#![doc(
html_logo_url = "https://raw.githubusercontent.com/neuronika/neuronika/main/misc/neuronika_brain.svg"
)]
#![doc(
html_favicon_url = "https://raw.githubusercontent.com/neuronika/neuronika/main/misc/neuronika_brain.ico"
)]
pub mod data;
pub mod nn;
pub mod optim;
mod variable;
use ndarray::{Array, Array2, Dimension, Ix1, Ix2, ShapeBuilder};
use ndarray_rand::rand_distr::Uniform;
use ndarray_rand::RandomExt;
pub use variable::{
Backward, Cat, Convolve, ConvolveWithGroups, Data, Eval, Forward, Gradient, MatMatMul,
MatMatMulT, Overwrite, Param, Stack, Var, VarDiff, VecMatMul, VecVecMul,
};
use variable::{Input, InputBackward};
/// Creates a variable from a **[ndarray]** array that owns its data.
///
/// # Examples
///
/// ```
/// use ndarray;
/// use neuronika;
///
/// let a = ndarray::array![[1., 2.], [3.,4.]];
/// let t = neuronika::from_ndarray(a.clone());
///
/// assert_eq!(*t.data(), a);
/// ```
pub fn from_ndarray<D: Dimension>(array: Array<f32, D>) -> Var<Input<D>> {
Input::new(array)
}
/// Creates a variable with zeroed data.
///
/// The shape is of type [`ndarray::ShapeBuilder`].
///
/// # Examples
///
/// ```
/// use neuronika;
/// let t1 = neuronika::zeros(1);
/// let t2 = neuronika::zeros((1, 5));
/// let t3 = neuronika::zeros([1, 2, 3]);
///
/// assert_eq!(t1.data().shape(), &[1]);
/// assert_eq!(t2.data().shape(), &[1, 5]);
/// assert_eq!(t3.data().shape(), &[1, 2, 3]);
/// ```
pub fn zeros<D: Dimension, Sh: ShapeBuilder<Dim = D>>(shape: Sh) -> Var<Input<D>> {
Input::new(Array::from_elem(shape, 0.0))
}
/// Creates a variable with data filled with ones.
///
/// The shape is of type [`ndarray::ShapeBuilder`].
///
/// # Examples
///
/// ```
/// use neuronika;
/// let t1 = neuronika::ones(1);
/// let t2 = neuronika::ones((1, 5));
/// let t3 = neuronika::ones([1, 2, 3]);
///
/// assert_eq!(t1.data().shape(), &[1]);
/// assert_eq!(t2.data().shape(), &[1, 5]);
/// assert_eq!(t3.data().shape(), &[1, 2, 3]);
/// ```
pub fn ones<D: Dimension, Sh: ShapeBuilder<Dim = D>>(shape: Sh) -> Var<Input<D>> {
Input::new(Array::from_elem(shape, 1.0))
}
/// Creates a variable with data filled with a constant value.
///
/// `el` must be `f32` and the shape of type [`ndarray::ShapeBuilder`].
///
/// # Examples
///
/// ```
/// use neuronika;
/// let t1 = neuronika::full(1, 5.); // Filled with 5.0
/// let t2 = neuronika::full((1, 5), 6.); // Filled with 6.0
/// let t3 = neuronika::full([1, 2, 3], 8.); // Filled with 8.0
///
/// assert_eq!(t1.data().shape(), &[1]);
/// assert_eq!(t2.data().shape(), &[1, 5]);
/// assert_eq!(t3.data().shape(), &[1, 2, 3]);
/// ```
pub fn full<D: Dimension, Sh: ShapeBuilder<Dim = D>>(shape: Sh, elem: f32) -> Var<Input<D>> {
Input::new(Array::from_elem(shape, elem))
}
/// Creates a variable with values sampled from a uniform distribution on the interval *[0,1)*.
///
/// The shape is of type [`ndarray::ShapeBuilder`].
///
/// # Examples
///
/// ```
/// use neuronika;
/// let t = neuronika::rand([4, 5, 6]);
///
/// assert_eq!(t.data().shape(), &[4, 5, 6]);
/// ```
pub fn rand<D: Dimension, Sh: ShapeBuilder<Dim = D>>(shape: Sh) -> Var<Input<D>> {
Input::new(Array::random(shape, Uniform::new(0., 1.)))
}
/// Creates a variable with an identity matrix of size *n*.
///
/// # Panics
///
/// If `n * n` would overflow `isize`.
///
/// # Examples
///
/// ```
/// use neuronika;
/// use ndarray::Array2;
///
/// let tensor = neuronika::eye(3);
/// assert_eq!(*tensor.data(), Array2::eye(3));
/// ```
pub fn eye(n: usize) -> Var<Input<Ix2>> {
Input::new(Array2::eye(n))
}
/// Creates a one-dimensional variable with *n* evenly spaced elements.
///
/// The elements range from `start` to `end` (exclusive).
///
/// # Panics
///
/// If the length is greater than [`isize::MAX`].
///
/// [`isize::MAX`]: https://doc.rust-lang.org/std/primitive.isize.html#associatedconstant.MAX
///
/// # Examples
///
/// ```
/// use neuronika;
/// use ndarray::arr1;
///
/// let tensor = neuronika::linspace(0., 1., 5);
/// assert!(*tensor.data() == arr1(&[0.0, 0.25, 0.5, 0.75, 1.0]))
/// ```
pub fn linspace(start: f32, end: f32, n: usize) -> Var<Input<Ix1>> {
Input::new(Array::linspace(start, end, n))
}
/// Creates a one-dimensional variable with *n* logarithmically spaced elements.
///
/// The starting value is `base.powf(start)` and the final one is `base.powf(end)`.
///
/// If `base` is negative, all values will be negative.
///
/// # Panics
///
/// If `n` is greater than [`isize::MAX`] or if converting `n - 1` to type `f32` fails.
///
/// [`isize::MAX`]: https://doc.rust-lang.org/std/primitive.isize.html#associatedconstant.MAX
pub fn logspace(base: f32, start: f32, end: f32, n: usize) -> Var<Input<Ix1>> {
Input::new(Array::logspace(base, start, end, n))
}
/// Creates a one-dimensional variable with *n* geometrically spaced elements.
///
/// The elements range from `start` to `end` (inclusive).
///
/// Returns `None` if `start` and `end` have different signs or if either one is zero. Conceptually,
/// this means that in order to obtain a `Some` result, `end / start` must be positive.
///
/// # Panics
///
/// If `n` is greater than [`isize::MAX`] or if converting `n - 1` to type `f32` fails.
///
/// [`isize::MAX`]: https://doc.rust-lang.org/std/primitive.isize.html#associatedconstant.MAX
pub fn geomspace(start: f32, end: f32, n: usize) -> Option<Var<Input<Ix1>>> {
Array::geomspace(start, end, n).map(Input::new)
}
/// Creates a one-dimensional variable with elements from *start* to *end* spaced by *step*.
///
/// The elements range from `start` to `end` (exclusive).
///
/// # Panics
///
/// If the length is greater than
/// [`isize::MAX`].
///
/// [`isize::Max`]: https://doc.rust-lang.org/std/primitive.isize.html#associatedconstant.MAX
///
/// # Examples
///
/// ```
/// use neuronika;
/// use ndarray::arr1;
///
/// let tensor = neuronika::range(0., 5., 1.);
/// assert!(*tensor.data() == arr1(&[0., 1., 2., 3., 4.]))
/// ```
pub fn range(start: f32, end: f32, step: f32) -> Var<Input<Ix1>> {
Input::new(Array::range(start, end, step))
}
/// Concatenates the variables `lhs` and `rhs` along `axis`.
///
/// All variables must have the same shape, except in the concatenating dimension.
///
/// # Arguments
///
/// * `lhs` - variable.
///
/// * `rhs` - other variable.
///
/// * `axis` - axis to concatenate along to.
///
/// # Panics
///
/// If the variables have mismatching shapes, apart from along axis, if the variables are empty,
/// if `axis` is out of bounds or if the result is larger than is possible to represent.
pub fn cat<Lhs, Rhs>(lhs: Lhs, rhs: Rhs, axis: usize) -> <Lhs as Cat<Rhs>>::Output
where
Lhs: Cat<Rhs>,
{
Cat::cat(lhs, rhs, axis)
}
/// Stacks the variables `lhs` and `rhs` along `axis`.
///
/// All variables must have the same shape.
///
/// # Arguments
///
/// * `lhs` - variable.
///
/// * `rhs` - other variable.
///
/// * `axis` - axis to stack along to.
///
/// # Panics
///
/// If the variables have mismatching shapes, apart from along axis, if the variables are empty,
/// if `axis` is out of bounds or if the result is larger than is possible to represent.
pub fn stack<Lhs, Rhs>(lhs: Lhs, rhs: Rhs, axis: usize) -> <Lhs as Stack<Rhs>>::Output
where
Lhs: Stack<Rhs>,
{
Stack::stack(lhs, rhs, axis)
}
#[cfg(test)]
mod tests {
#[test]
fn from_ndarray_test() {
use super::from_ndarray;
let a = ndarray::array![[1., 2.], [3., 4.]];
let t = from_ndarray(a.clone());
assert_eq!(*t.data(), a);
}
#[test]
fn zeros() {
use super::zeros;
let t1 = zeros(1);
let t2 = zeros((1, 5));
let t3 = zeros([1, 2, 3]);
assert_eq!(t1.data().shape(), &[1]);
assert_eq!(t2.data().shape(), &[1, 5]);
assert_eq!(t3.data().shape(), &[1, 2, 3]);
assert!(
t1.data().iter().all(|el| *el <= f32::EPSILON)
&& t2.data().iter().all(|el| *el <= f32::EPSILON)
&& t3.data().iter().all(|el| *el <= f32::EPSILON)
)
}
#[test]
fn ones() {
use super::ones;
let t1 = ones(1);
let t2 = ones((1, 5));
let t3 = ones([1, 2, 3]);
assert_eq!(t1.data().shape(), &[1]);
assert_eq!(t2.data().shape(), &[1, 5]);
assert_eq!(t3.data().shape(), &[1, 2, 3]);
assert!(
t1.data().iter().all(|el| (*el - 1.).abs() <= f32::EPSILON)
&& t2.data().iter().all(|el| (*el - 1.).abs() <= f32::EPSILON)
&& t3.data().iter().all(|el| (*el - 1.).abs() <= f32::EPSILON)
)
}
#[test]
fn full() {
use super::full;
let t1 = full(1, 5.);
let t2 = full((1, 5), 6.);
let t3 = full([1, 2, 3], 8.);
assert!(
t1.data().iter().all(|el| (*el - 5.).abs() <= f32::EPSILON)
&& t2.data().iter().all(|el| (*el - 6.).abs() <= f32::EPSILON)
&& t3.data().iter().all(|el| (*el - 8.).abs() <= f32::EPSILON)
)
}
#[test]
fn rand_test() {
use super::rand;
let t = rand([4, 5, 6]);
assert_eq!(t.data().shape(), &[4, 5, 6]);
}
#[test]
fn eye_test() {
use super::{eye, Array2};
let tensor = eye(3);
assert_eq!(*tensor.data(), Array2::<f32>::eye(3));
}
#[test]
fn linspace() {
use super::linspace;
let tensor = linspace(0., 1., 5);
assert!(*tensor.data() == ndarray::arr1(&[0.0, 0.25, 0.5, 0.75, 1.0]))
}
#[test]
fn logspace() {
use super::logspace;
let tensor = logspace(2., 1., 5., 5);
assert!(*tensor.data() == ndarray::arr1(&[2., 4., 8., 16., 32.]))
}
#[test]
fn geomspace() {
use super::geomspace;
let tensor = geomspace(1., 1000., 4);
assert!(tensor
.unwrap()
.data()
.iter()
.zip(ndarray::arr1(&[1.0_f32, 10.0_f32, 100.0_f32, 1000.0_f32]).iter())
.all(|(&t, &a)| (t.round() - a.round()).abs() <= f32::EPSILON));
}
#[test]
fn range_test() {
use super::*;
let tensor = range(0., 5., 1.);
assert!(*tensor.data() == ndarray::arr1(&[0., 1., 2., 3., 4.]))
}
}
#[test]
fn kcckk() {
let n = crate::from_ndarray(ndarray::array![1., 2.]);
println!("{}", n);
}