Skip to content

Additional show info for PCA #90

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
wildart merged 13 commits into from
Mar 3, 2022
Merged

Additional show info for PCA #90

Show file tree
Hide file tree
Changes from 4 commits
Commits
Show all changes
13 commits
Select commit Hold shift + click to select a range
923ca2f
Additional `show` info for PCA
Tokazama Apr 7, 2019
5c9b6dd
Add show method
Tokazama Sep 4, 2019
6540e8f
Added CoefTable for pretty printing.
Sep 4, 2019
78c24ae
Fixed row orientation
Sep 4, 2019
d313dd4
Update src/pca.jl
Tokazama Sep 5, 2019
2cb0251
Update src/pca.jl
Tokazama Sep 5, 2019
e31458a
Update src/pca.jl
Tokazama Sep 5, 2019
07331d6
Update src/pca.jl
Tokazama Sep 5, 2019
03bdfd4
Merge branch 'master' into master
wildart Oct 28, 2021
63baa5a
Updated output tables
wildart Feb 17, 2022
16c9b75
Merge branch 'master' into master
wildart Feb 17, 2022
15f96d4
Update src/pca.jl
Tokazama Feb 17, 2022
6807a4a
Update src/pca.jl
Tokazama Feb 17, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/MultivariateStats.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,7 +3,7 @@ module MultivariateStats
using StatsBase: SimpleCovariance, CovarianceEstimator
import Statistics: mean, var, cov, covm
import Base: length, size, show, dump
import StatsBase: fit, predict, ConvergenceException
import StatsBase: fit, predict, ConvergenceException, CoefTable

@nalimilan nalimilan Sep 5, 2019

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

import is for when you overload a function with a new method. It shouldn't be needed here.

import SparseArrays

export
Expand Down
18 changes: 17 additions & 1 deletion src/pca.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -47,11 +47,27 @@ transform(M::PCA{T}, x::AbstractVecOrMat{T}) where {T<:Real} = transpose(M.proj)
reconstruct(M::PCA{T}, y::AbstractVecOrMat{T}) where {T<:Real} = decentralize(M.proj * y, M.mean)

## show & dump

function show(io::IO, M::PCA)
print(io, "PCA(indim = $(indim(M)), outdim = $(outdim(M)), principalratio = $(principalratio(M)))")
end

function show(io::IO, ::MIME"text/plain", M::PCA)
print(io, "PCA(indim = $(indim(M)), outdim = $(outdim(M)), principalratio = $(principalratio(M)))")
ldgs = projection(M) * diagm(0 => sqrt.(M.prinvars))
rot = diag(ldgs' * ldgs)
ldgs = ldgs[:,sortperm(rot, rev=true)]
Comment thread
Tokazama marked this conversation as resolved.
Outdated
ldgs_signs = sign.(sum(ldgs, dims=1))
ldgs_signs[ldgs_signs .== 0] .= 1
Comment thread
Tokazama marked this conversation as resolved.
Outdated
ldgs = ldgs * diagm(0 => ldgs_signs[:])
print(io, "\n\nPattern matrix\n")

@nalimilan nalimilan Sep 5, 2019

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
print(io, "\n\nPattern matrix\n")
print(io, "\n\nPattern matrix:\n")

(for consistency with below)

Also, better be specific about what this matrix is: psych says "standardized loadings".

display(ldgs)

@Tokazama Tokazama Sep 4, 2019

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It would be nice if this came out with columns and rows labeled but without access to the variable names that's not possible. In the current state the columns are the components in the same order as those in the following coeftable.

@nalimilan nalimilan Sep 5, 2019

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes, that's annoying. For now maybe print a CoefTable object without row names. If you pass an empty vector as row names, numbers will automatically be generated. That way at least the names of PCs will be consistent with the table below.

Also, how about printing the second table first? That's the most important result I think.

Comment thread
Tokazama marked this conversation as resolved.
Outdated
print(io, "\n")
print(io, "Importance of components:\n")
print(io, CoefTable(vcat(principalvars(M)', (principalvars(M) ./ tvar(M))', (cumsum(principalvars(M) ./tvar(M)))'),
string.("PC", 1:length(principalvars(M))), # components in order
["Loadings", "Proportion explained", "Cumulative proportion"])) # row names

@nalimilan nalimilan Sep 5, 2019

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

"Loadings" is "SS loadings" from psych, right? Shouldn't that be specified? Also "proportion" should say of what (the variance). And how about adding "proportion of variance" and "cumulative variance" which are in base R and psych?

@wildart wildart Sep 5, 2019

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

They are not "Loadings", they are "Eigenvalues" or "SS loadings" as R calls them.

Loadings would be

projection(M).*sqrt.(principalvars(M))'

@Tokazama Tokazama Sep 5, 2019

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@wildart You're definitely right. I apologize for the oversight.

end
Comment thread
Tokazama marked this conversation as resolved.

function dump(io::IO, M::PCA)
show(io, M)
println(io)
Expand Down