FIR-Scaled duration predictors

using Unfold
using Interpolations

using UnfoldSim
using UnfoldMakie, CairoMakie
using DataFrames

data, evts = UnfoldSim.predef_eeg(sfreq = 10, n_repeats = 1)

evts.duration = 5:24
5:24

putting scale_duration = Interpolation.Linear() will introduce a Cameron-Hassall 2022 PNAS- Style basisfunction, that scales with the :duration column

basisfunction = firbasis(ฯ„ = (-1, 2), sfreq = 5, scale_duration = Interpolations.Linear())
โ•ญโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ•ฎ8ร—16 SparseArrays.SparseMatrixCSC with 16 stored entries:
โ”‚  ::BasisFunction                                                             โ”‚โŽกโ ‰โ ’โ คโฃ€โ €โ €โ €โ €โŽค                                               
โ”‚  name:                                                                       โ”‚โŽฃโ €โ €โ €โ €โ ‰โ ’โ คโฃ€โŽฆ                                               
โ”‚  kerneltype: FIRBasis                                                        โ”‚                                                         
โ”‚  width: 16                                                                   โ”‚                                                         
โ”‚  height: 0                                                                   โ”‚                                                         
โ”‚  colnames: [-1.0, -0.8 ... 2.0]                                              โ”‚                                                         
โ”‚  times: [-1.0, -0.8 ... 2.0]                                                 โ”‚                                                         
โ”‚  collabel: time                                                              โ”‚                                                         
โ”‚  shift_onset: -5                                                             โ”‚                                                         
โ”‚                                                                              โ”‚                                                         
โ•ฐโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ•ฏ                                                         

Two examples with duration = 10

Unfold.kernel(basisfunction, [0, 10])
10ร—16 SparseArrays.SparseMatrixCSC{Float64, Int64} with 20 stored entries:
โŽกโ ™โ ขโข„โก€โ €โ €โ €โ €โŽค
โŽขโ €โ €โ €โ ‰โ ฒโข„โก€โ €โŽฅ
โŽฃโ €โ €โ €โ €โ €โ €โ ˆโ “โŽฆ

and duration = 20

Unfold.kernel(basisfunction, [0, 20])
20ร—16 SparseArrays.SparseMatrixCSC{Float64, Int64} with 38 stored entries:
โŽกโ ณโฃ„โ €โ €โ €โ €โ €โ €โŽค
โŽขโ €โ ˜โขงโก€โ €โ €โ €โ €โŽฅ
โŽขโ €โ €โ €โ นโฃ†โ €โ €โ €โŽฅ
โŽขโ €โ €โ €โ €โ ˆโขณโก„โ €โŽฅ
โŽฃโ €โ €โ €โ €โ €โ €โ ™โขฆโŽฆ

let's fit a model

f = @formula 0 ~ 1 + condition
bf_vec = [Any => (f, basisfunction)]
m = fit(UnfoldModel, bf_vec, evts, data; eventfields = [:latency, :duration]);


# currently bugged for small matrices
# plot_designmatrix(designmatrix(m))
# thus using
heatmap(Matrix(modelmatrix(m))')
Example block output

As one can see, the designmatrix is nicely scaled

We can predict overlap-corrected results

p = predict(m; overlap = false)[1]
heatmap(p[1, :, :])
Example block output

note the missings which are displayed as white pixels.

Block-design predictors

In contrast, it is also possible to put scale_duration = true - which wil not scale the matrix as before, but introduce a step-function.

putting scale_duration = Interpolation.Linear() will introduce a Cameron-Hassall 2022 PNAS- Style basisfunction, that scales with the :duration column

basisfunction = firbasis(ฯ„ = (-1, 2), sfreq = 5, scale_duration = true)
โ•ญโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ•ฎ23ร—16 SparseArrays.SparseMatrixCSC with 128 stored entries:
โ”‚  ::BasisFunction                                                             โ”‚โŽกโฃทโฃ„โ €โ €โ €โ €โ €โ €โŽค                                                 
โ”‚  name:                                                                       โ”‚โŽขโฃฟโฃฟโฃทโฃ„โ €โ €โ €โ €โŽฅ                                                 
โ”‚  kerneltype: FIRBasis                                                        โ”‚โŽขโ ˆโ ปโฃฟโฃฟโฃทโฃ„โ €โ €โŽฅ                                                 
โ”‚  width: 16                                                                   โ”‚โŽขโ €โ €โ ˆโ ปโฃฟโฃฟโฃทโฃ„โŽฅ                                                 
โ”‚  height: 16                                                                  โ”‚โŽขโ €โ €โ €โ €โ ˆโ ปโฃฟโฃฟโŽฅ                                                 
โ”‚  colnames: [-1.0, -0.8 ... 2.0]                                              โ”‚โŽฃโ €โ €โ €โ €โ €โ €โ ˆโ ปโŽฆ                                                 
โ”‚  times: [-1.0, -0.8 ... 2.0]                                                 โ”‚                                                           
โ”‚  collabel: time                                                              โ”‚                                                           
โ”‚  shift_onset: -5                                                             โ”‚                                                           
โ”‚                                                                              โ”‚                                                           
โ•ฐโ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ”€โ•ฏ                                                           

Two examples with duration = 10

Unfold.kernel(basisfunction, [0, 10])
25ร—16 SparseArrays.SparseMatrixCSC{Int64, Int64} with 160 stored entries:
โŽกโฃทโฃ„โ €โ €โ €โ €โ €โ €โŽค
โŽขโฃฟโฃฟโฃทโฃ„โ €โ €โ €โ €โŽฅ
โŽขโ ปโฃฟโฃฟโฃฟโฃทโฃ„โ €โ €โŽฅ
โŽขโ €โ ˆโ ปโฃฟโฃฟโฃฟโฃทโฃ„โŽฅ
โŽขโ €โ €โ €โ ˆโ ปโฃฟโฃฟโฃฟโŽฅ
โŽขโ €โ €โ €โ €โ €โ ˆโ ปโฃฟโŽฅ
โŽฃโ €โ €โ €โ €โ €โ €โ €โ ˆโŽฆ

and duration = 20

Unfold.kernel(basisfunction, [0, 20])
35ร—16 SparseArrays.SparseMatrixCSC{Int64, Int64} with 320 stored entries:
โŽกโฃทโฃ„โ €โ €โ €โ €โ €โ €โŽค
โŽขโฃฟโฃฟโฃทโฃ„โ €โ €โ €โ €โŽฅ
โŽขโฃฟโฃฟโฃฟโฃฟโฃทโฃ„โ €โ €โŽฅ
โŽขโฃฟโฃฟโฃฟโฃฟโฃฟโฃฟโฃทโฃ„โŽฅ
โŽขโฃฟโฃฟโฃฟโฃฟโฃฟโฃฟโฃฟโฃฟโŽฅ
โŽขโ ˆโ ปโฃฟโฃฟโฃฟโฃฟโฃฟโฃฟโŽฅ
โŽขโ €โ €โ ˆโ ปโฃฟโฃฟโฃฟโฃฟโŽฅ
โŽขโ €โ €โ €โ €โ ˆโ ปโฃฟโฃฟโŽฅ
โŽฃโ €โ €โ €โ €โ €โ €โ ˆโ ปโŽฆ

let's fit a model

f = @formula 0 ~ 1 + condition
bf_vec = [Any => (f, basisfunction)]
m = fit(UnfoldModel, bf_vec, evts, data; eventfields = [:latency, :duration]);


heatmap(Matrix(modelmatrix(m))')
Example block output

as one can see, now the designmatrix is not stretched - but rather "block"-ed

p = predict(m; overlap = false)[1]
heatmap(p[1, :, :])
Example block output

This page was generated using Literate.jl.