Skip to content

Commit

Permalink
Update docs
Browse files Browse the repository at this point in the history
  • Loading branch information
dalcde committed Jan 24, 2022
1 parent 76d11dc commit 0f19a29
Show file tree
Hide file tree
Showing 7 changed files with 47 additions and 14 deletions.
2 changes: 1 addition & 1 deletion ext/crates/fp/src/matrix/matrix_inner.rs
Original file line number Diff line number Diff line change
Expand Up @@ -1146,7 +1146,7 @@ impl<'a> MatrixSliceMut<'a> {
}
}

/// For each row, add the v[i]th entry of `other` to `self`.
/// For each row, add the `v[i]`th entry of `other` to `self`.
pub fn add_masked(&mut self, other: &Matrix, mask: &[usize]) {
assert_eq!(self.rows(), other.rows());

Expand Down
2 changes: 1 addition & 1 deletion ext/examples/resolve_through_stem.rs
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
//! Resolves a module up to a $(s, n)$ and prints an ASCII depiction of the Ext groups:
//! Resolves a module up to an $(n, s)$ and prints an ASCII depiction of the Ext groups:
//! ```text
//! ·
//! · ·
Expand Down
11 changes: 3 additions & 8 deletions ext/examples/secondary.rs
Original file line number Diff line number Diff line change
@@ -1,17 +1,12 @@
//! This computes $d_2$ differentials in the Adams spectral sequence. This only works for fairly
//! specific modules, but tends to cover most cases of interest.
//!
//! In general, the set of possible $d_2$'s is a torsor over $\Ext^{2, 1}(M, M)$; the
//! action of $\chi \in \Ext^{2, 1}(M, M)$ is given by adding $\chi$-multiplication
//! to the $d_2$ map. This algorithm computes one possible set of $d_2$'s. If $\Ext^{2, 1}(M, M)$
//! is non-zero, some differentials will have to be calculated by hand to determine the actual set
//! of $d_2$'s.
//!
//! # Usage
//! This asks for a module in the usual way. It only works with the Milnor basis,
//! This asks for a module in the usual way, and verifies that the module satisfies the conditions
//! necessary for the algorithm the work. It only works with the Milnor basis.
//!
//! # Output
//! We omit differentials if the target bidegree is zero
//! We omit differentials if the target bidegree is zero.
use algebra::module::Module;
use std::sync::Arc;
Expand Down
24 changes: 23 additions & 1 deletion ext/examples/secondary_massey.rs
Original file line number Diff line number Diff line change
@@ -1,4 +1,26 @@
//! Computes massey products in $\Mod_{C\tau^2}$. This is only correct up to a sign!!!
//! Computes massey products in $\Mod_{C\tau^2}$.
//!
//! # Usage
//! This computes all Massey products of the form $\langle a, b, -\rangle$, where
//! $a \in \Ext^{\*, \*}(M, k)$ and $b, (-) \in \Ext^{\*, \*}(k, k)$. It does not verify that the
//! Massey product is valid, i.e. $a$ and $b$ both lift to $\Mod_{C\tau^2}$ and have trivial
//! product.
//!
//! Since we must choose $a$ and $b$ to have trivial product, it is necessary to be able to specify
//! the $\tau$ part of them, and not insist that they are standard lifts of the $\Ext$ classes.
//! Thus, the user is first prompted for the $\Ext$ part, then the $\tau$ part of each class. If
//! the bidegree right above the class is empty, the user is not prompted for the $\tau$ part.
//!
//! Note that for the purposes of save files, the name of the product should include the $\tau$
//! part as well. Products with different $\tau$ parts should have separate save directories. (If
//! one wishes to save some of computation, one can symlink the `secondary_composites` directory,
//! which *can* be shared for products with the same $\Ext$ part but different $\tau$ part) There
//! is no theoretical reason this has to be the case, but it is rather unpleasant to make it use
//! the same save file, and such a need has not arisen yet.
//!
//! # Output
//! This computes the Massey products up to a sign. Brave souls are encouraged to figure out the
//! correct sign for the products.
use std::path::PathBuf;
use std::sync::Arc;
Expand Down
16 changes: 16 additions & 0 deletions ext/examples/secondary_product.rs
Original file line number Diff line number Diff line change
@@ -1,4 +1,20 @@
//! Computes products in $\Mod_{C\tau^2}$.
//!
//! # Usage
//! The program asks for a module $M$ and an element $x \in \Ext^{\*, \*}(M, k)$. It then
//! computes the secondary product of the standard lift of $x$ with all (standard lifts of)
//! elements in $\Ext^{\*, \*}(M, k)$ that survive $d_2$.
//!
//! These products are computed for all elements whose product with $x$ lies in the specified
//! bidegree of $M$, and $k$ is resolved as far as necessary to support this computation.
//!
//! Running this program requires computing the secondary resolution of both $M$ and $k$, i.e. the
//! calculations performed by [`secondary`](../secondary/index.html). The user is encouraged to
//! make use of a save file to reuse these calculations for different products. (When $M$ is not
//! equal to $k$, the user will be prompted for the save directory of $k$)
//!
//! # Notes
//! The program verifies that $x$ is indeed permanent.
use std::path::PathBuf;
use std::sync::Arc;
Expand Down
4 changes: 2 additions & 2 deletions ext/src/lib.rs
Original file line number Diff line number Diff line change
Expand Up @@ -159,8 +159,8 @@
//! configured via the `RAYON_NUM_THREADS` environment variable.
//! - `use-zstd`: Support reading zstd-compressed save files.
//! - `cache-multiplication`: Precompute and cache the multiplication table under the Milnor basis.
//! This is only feasible when using a small, finite subalgebra, e.g. when working with tmf
//! modules.
//! This is only feasible when using a small, finite subalgebra, e.g. when working with
//! $\mathrm{tmf}$ modules.
#![allow(clippy::upper_case_acronyms)]

Expand Down
2 changes: 1 addition & 1 deletion ext/src/nassau.rs
Original file line number Diff line number Diff line change
Expand Up @@ -17,8 +17,8 @@ use once::OnceVec;
#[cfg(feature = "concurrent")]
use std::sync::mpsc;

/// See [`Resolution::SenderData`]. This differs by not having the `new` field.
#[cfg(feature = "concurrent")]
/// See [`resolution::SenderData`](../resolution/struct.SenderData.html). This differs by not having the `new` field.
struct SenderData {
s: u32,
t: i32,
Expand Down

0 comments on commit 0f19a29

Please sign in to comment.