From 86ad2012fd3e9443c9799b5b46e585224253beae Mon Sep 17 00:00:00 2001 From: Oren Ben-Kiki Date: Sat, 1 Jun 2024 18:33:21 +0300 Subject: [PATCH] Factor out transpose!. --- docs/v0.1.0/.documenter-siteinfo.json | 2 +- docs/v0.1.0/index.html | 18 ++- docs/v0.1.0/matrix_layouts.html | 99 +++++++++++++++- docs/v0.1.0/objects.inv | Bin 4164 -> 4194 bytes docs/v0.1.0/read_only.html | 64 ----------- docs/v0.1.0/search_index.js | 2 +- src/matrix_layouts.jl | 158 +++++++++++++++++++++----- src/matrix_layouts.md | 2 + src/operations.jl | 2 +- src/queries.jl | 1 + src/read_only.jl | 83 +------------- src/read_only.md | 1 - test/matrix_layouts.jl | 20 ++++ 13 files changed, 267 insertions(+), 185 deletions(-) diff --git a/docs/v0.1.0/.documenter-siteinfo.json b/docs/v0.1.0/.documenter-siteinfo.json index 00ffee0..7a47f83 100644 --- a/docs/v0.1.0/.documenter-siteinfo.json +++ b/docs/v0.1.0/.documenter-siteinfo.json @@ -1 +1 @@ -{"documenter":{"julia_version":"1.10.0","generation_timestamp":"2024-06-01T14:35:03","documenter_version":"1.4.1"}} \ No newline at end of file +{"documenter":{"julia_version":"1.10.0","generation_timestamp":"2024-06-01T18:28:22","documenter_version":"1.4.1"}} \ No newline at end of file diff --git a/docs/v0.1.0/index.html b/docs/v0.1.0/index.html index 5ba150b..d8cd432 100644 --- a/docs/v0.1.0/index.html +++ b/docs/v0.1.0/index.html @@ -1712,6 +1712,12 @@

  • + +Daf.MatrixLayouts.copy_array + + +
    • +
  • Daf.MatrixLayouts.depict @@ -1868,12 +1874,6 @@

  • - -Daf.ReadOnly.copy_array - - -
    • -
  • Daf.ReadOnly.is_read_only_array @@ -2162,6 +2162,12 @@

  • + +LinearAlgebra.transpose! + + +
    • +
  • Daf.Computations.@computation diff --git a/docs/v0.1.0/matrix_layouts.html b/docs/v0.1.0/matrix_layouts.html index 290dd3b..a132f0b 100644 --- a/docs/v0.1.0/matrix_layouts.html +++ b/docs/v0.1.0/matrix_layouts.html @@ -576,8 +576,10 @@

    , summing the UMIs of a gene is fast, and summing the UMIs of a cell is slow.

    In contrast, + transpose! - (with a + + (with a ! ) is slow; it creates a rearranged copy of the data, also returning a matrix whose rows are genes and columns are cells, but this time, in column-major layout. Therefore, in this case summing the UMIs of a gene will be slow, and summing the UMIs of a cell will be fast.

    @@ -650,6 +652,89 @@

    +
    +
    + + + +LinearAlgebra.transpose! + + — +Function + +
    +
    +
    +
    +transpose!(matrix::AbstractMatrix)::AbstractMatrix
+transpose!(matrix::NamedMatrix)::NamedMatrix
+
+
    +

    This is a shorthand for +LinearAlgebra.transpose!(similar(transpose(m)), m) +. That is, this will return a transpose of a matrix, but instead of simply using a zero-copy wrapper, it actually rearranges the data. See + +relayout! + +. +

    +
    +
    +
    +
    +
    + + + +Daf.MatrixLayouts.copy_array + + — +Function + +
    +
    +
    +
    +copy_array(array::AbstractArray)::AbstractArray
+
+
    +

    Create a mutable copy of an array. This differs from +Base.copy + in the following: +

    +
      +
    • Copying a read-only array is a mutable array. In contrast, both +Base.copy + and +Base.deepcopy + of a read-only array will return a read-only array, which is technically correct, but is rather pointless for +Base.copy +. +
      • +
    • Copying will preserve the layout of the data; for example, copying a +Transpose + array is still a +Transpose + array. In contrast, while +Base.deepcopy + will preserve the layout, +Base.copy + will silently + +relayout! + + the matrix, which is both expensive and confusing. +
      • +
    • Copying a sparse vector or matrix gives the same type of sparse array or matrix. Copying anything else gives a simple dense array regardless of the original type. This is done because a +deepcopy + of +PyArray + will still share the underlying buffer. Sigh. +
      • +
    +
    +
    +

    Ensuring code efficiency @@ -789,6 +874,12 @@

  • + +Daf.MatrixLayouts.copy_array + + +
    • +
  • Daf.MatrixLayouts.inefficient_action_handler @@ -830,6 +921,12 @@

    • +
  • + +LinearAlgebra.transpose! + + +