-
Notifications
You must be signed in to change notification settings - Fork 1
/
README.md
253 lines (191 loc) · 8.41 KB
/
README.md
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
<!-- README.md is generated from README.Rmd. Please edit that file -->
# jlmerclusterperm <a href="https://yjunechoe.github.io/jlmerclusterperm/"><img src="man/figures/logo.png" align="right" height="150" /></a>
<!-- badges: start -->
[![CRAN
status](https://www.r-pkg.org/badges/version/jlmerclusterperm)](https://CRAN.R-project.org/package=jlmerclusterperm)
[![R-CMD-check](https://github.com/yjunechoe/jlmerclusterperm/actions/workflows/R-CMD-check.yaml/badge.svg)](https://github.com/yjunechoe/jlmerclusterperm/actions/workflows/R-CMD-check.yaml)
[![pkgcheck](https://github.com/yjunechoe/jlmerclusterperm/workflows/pkgcheck/badge.svg)](https://github.com/yjunechoe/jlmerclusterperm/actions?query=workflow%3Apkgcheck)
[![Codecov test
coverage](https://codecov.io/gh/yjunechoe/jlmerclusterperm/branch/main/graph/badge.svg)](https://app.codecov.io/gh/yjunechoe/jlmerclusterperm?branch=main)
[![CRAN
downloads](https://cranlogs.r-pkg.org/badges/grand-total/jlmerclusterperm)](https://cranlogs.r-pkg.org/badges/grand-total/jlmerclusterperm)
<!-- badges: end -->
Julia [GLM.jl](https://github.com/JuliaStats/GLM.jl) and
[MixedModels.jl](https://github.com/JuliaStats/MixedModels.jl) based
implementation of the cluster-based permutation test for time series
data, powered by
[JuliaConnectoR](https://github.com/stefan-m-lenz/JuliaConnectoR).
![](man/figures/clusterpermute_slice.png)
## Installation and usage
Install the released version of jlmerclusterperm from CRAN:
``` r
install.packages("jlmerclusterperm")
```
Or install the development version from
[GitHub](https://github.com/yjunechoe/jlmerclusterperm) with:
``` r
# install.packages("remotes")
remotes::install_github("yjunechoe/jlmerclusterperm")
```
Using `jlmerclusterperm` requires a prior installation of the Julia
programming language, which can be downloaded from either the [official
website](https://julialang.org/) or using the command line utility
[juliaup](https://github.com/JuliaLang/juliaup). Julia version \>=1.8 is
required and
[1.9](https://julialang.org/blog/2023/04/julia-1.9-highlights/#caching_of_native_code)
or higher is preferred for the substantial speed improvements.
Before using functions from `jlmerclusterperm`, an initial setup is
required via calling `jlmerclusterperm_setup()`. The very first call on
a system will install necessary dependencies (this only happens once and
takes around 10-15 minutes).
Subsequent calls to `jlmerclusterperm_setup()` incur a small overhead of
around 30 seconds, plus slight delays for first-time function calls. You
pay up front for start-up and warm-up costs and get blazingly-fast
functions from the package.
``` r
# Both lines must be run at the start of each new session
library(jlmerclusterperm)
jlmerclusterperm_setup()
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/setup-io-dark.svg">
<img src="man/figures/README-/setup-io.svg" style="display: block; margin: auto;" />
</picture>
See the [Get
Started](https://yjunechoe.github.io/jlmerclusterperm/articles/jlmerclusterperm.html)
page on the [package
website](https://yjunechoe.github.io/jlmerclusterperm/) for background
and tutorials.
## Quick tour of package functionalities
### Wholesale CPA with `clusterpermute()`
A time series data:
``` r
chickweights <- ChickWeight
chickweights$Time <- as.integer(factor(chickweights$Time))
matplot(
tapply(chickweights$weight, chickweights[c("Time", "Diet")], mean),
type = "b", lwd = 3, ylab = "Weight", xlab = "Time"
)
```
<img src="man/figures/README-chickweight-1.png" width="75%" style="display: block; margin: auto;" />
Preparing a specification object with `make_jlmer_spec()`:
``` r
chickweights_spec <- make_jlmer_spec(
formula = weight ~ 1 + Diet,
data = chickweights,
subject = "Chick", time = "Time"
)
chickweights_spec
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/spec-io-dark.svg">
<img src="man/figures/README-/spec-io.svg" style="display: block; margin: auto;" />
</picture>
Cluster-based permutation test with `clusterpermute()`:
``` r
set_rng_state(123L)
clusterpermute(
chickweights_spec,
threshold = 2.5,
nsim = 100
)
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/CPA-io-dark.svg">
<img src="man/figures/README-/CPA-io.svg" style="display: block; margin: auto;" />
</picture>
Including random effects:
``` r
chickweights_re_spec <- make_jlmer_spec(
formula = weight ~ 1 + Diet + (1 | Chick),
data = chickweights,
subject = "Chick", time = "Time"
)
set_rng_state(123L)
clusterpermute(
chickweights_re_spec,
threshold = 2.5,
nsim = 100
)$empirical_clusters
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/reCPA-io-dark.svg">
<img src="man/figures/README-/reCPA-io.svg" style="display: block; margin: auto;" />
</picture>
### Piecemeal approach to CPA
Computing time-wise statistics of the observed data:
``` r
empirical_statistics <- compute_timewise_statistics(chickweights_spec)
matplot(t(empirical_statistics), type = "b", pch = 1, lwd = 3, ylab = "t-statistic")
abline(h = 2.5, lty = 3)
```
<img src="man/figures/README-empirical_statistics-1.png" width="75%" style="display: block; margin: auto;" />
Identifying empirical clusters:
``` r
empirical_clusters <- extract_empirical_clusters(empirical_statistics, threshold = 2.5)
empirical_clusters
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/empirical_clusters-dark.svg">
<img src="man/figures/README-/empirical_clusters.svg" style="display: block; margin: auto;" />
</picture>
Simulating the null distribution:
``` r
set_rng_state(123L)
null_statistics <- permute_timewise_statistics(chickweights_spec, nsim = 100)
null_cluster_dists <- extract_null_cluster_dists(null_statistics, threshold = 2.5)
null_cluster_dists
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/null_statistics-dark.svg">
<img src="man/figures/README-/null_statistics.svg" style="display: block; margin: auto;" />
</picture>
Significance testing the cluster-mass statistic:
``` r
calculate_clusters_pvalues(empirical_clusters, null_cluster_dists, add1 = TRUE)
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/calculate_clusters_pvalues-dark.svg">
<img src="man/figures/README-/calculate_clusters_pvalues.svg" style="display: block; margin: auto;" />
</picture>
Iterating over a range of threshold values:
``` r
walk_threshold_steps(empirical_statistics, null_statistics, steps = c(2, 2.5, 3))
```
<picture>
<source media="(prefers-color-scheme: dark)" srcset="man/figures/README-/walk_threshold_steps-dark.svg">
<img src="man/figures/README-/walk_threshold_steps.svg" style="display: block; margin: auto;" />
</picture>
## Acknowledgments
- The paper [Maris & Oostenveld
(2007)](https://doi.org/10.1016/j.jneumeth.2007.03.024) which
originally proposed the cluster-based permutation analysis.
- The [JuliaConnectoR](https://github.com/stefan-m-lenz/JuliaConnectoR)
package for powering the R interface to Julia.
- The Julia packages [GLM.jl](https://github.com/JuliaStats/GLM.jl) and
[MixedModels.jl](https://github.com/JuliaStats/MixedModels.jl) for
fast implementations of (mixed effects) regression models.
- Existing implementations of CPA in R
([permuco](https://jaromilfrossard.github.io/permuco/),
[permutes](https://cran.r-project.org/package=permutes), etc.) whose
designs inspired the CPA interface in jlmerclusterperm.
## Citations
If you use jlmerclusterperm for cluster-based permutation test with
mixed-effects models in your research, please cite one (or more) of the
following as you see fit.
To cite jlmerclusterperm:
- Choe, J. (2023). jlmerclusterperm: Cluster-Based Permutation Analysis
for Densely Sampled Time Data. R package version 1.1.2.
<https://cran.r-project.org/package=jlmerclusterperm>.
To cite the cluster-based permutation test:
- Maris, E., & Oostenveld, R. (2007). Nonparametric statistical testing
of EEG- and MEG-data. *Journal of Neuroscience Methods, 164*, 177–190.
doi: 10.1016/j.jneumeth.2007.03.024.
To cite the Julia programming language:
- Bezanson, J., Edelman, A., Karpinski, S., & Shah, V. B. (2017). Julia:
A Fresh Approach to Numerical Computing. *SIAM Review, 59*(1), 65–98.
doi: 10.1137/141000671.
To cite the GLM.jl and MixedModels.jl Julia libraries, consult their
Zenodo pages:
- GLM: <https://doi.org/10.5281/zenodo.3376013>
- MixedModels: <https://zenodo.org/badge/latestdoi/9106942>