SuSiE z Estimate Prior Variance Methods Comparison

Last updated: 2019-02-17

• ✔ R Markdown file: up-to-date

  Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

• ✔ Environment: empty

  Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

• ✔ Seed: set.seed(20190115)

  The command set.seed(20190115) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

• ✔ Session information: recorded

  Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

• ✔ Repository version: 1dbbe5a

  Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility. The version displayed above was the version of the Git repository at the time these results were generated.
  
  Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:

  
  Ignored files:
      Ignored:    .DS_Store
      Ignored:    .Rhistory
      Ignored:    .Rproj.user/
      Ignored:    .sos/
      Ignored:    data/.DS_Store
      Ignored:    output/.DS_Store
  
  Untracked files:
      Untracked:  docs/figure/test.Rmd/
      Untracked:  output/dsc_susie_z_v_output.rds
      Untracked:  output/dscoutProblem475.rds
      Untracked:  output/dscoutProblem75.rds
  
  

  Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

library(dscrutils)
dscout = dscquery('susie_z_v', target='sim_gaussian.pve sim_gaussian.n_signal sim_gaussian.effect_weight susie_z_uniroot.L susie_z_em.L susie_z_optim.L susie_z_uniroot.optimV_method susie_z_em.optimV_method susie_z_optim.optimV_method score_susie.objective score_susie.converged score_susie.total score_susie.valid susie_z_uniroot.DSC_TIME susie_z_em.DSC_TIME susie_z_optim.DSC_TIME')
colnames(dscout) = c('DSC', 'pve', 'n_signal', 'effect_weight', 'L_uniroot', 'method_uniroot','Time_uniroot', 'L_em', 'method_em', 'Time_em', 'L_optim', 'method_optim', 'Time_optim', 'objective', 'converged', 'total', 'valid')
dscout$effect_weight[which(dscout$effect_weight == 'rep(1/n_signal, n_signal)')] = 'equal'
dscout$effect_weight[which(dscout$effect_weight != 'equal')] = 'notequal'
method = dscout$method_uniroot
method[dscout$method_em == 'EM'] = 'em'
method[dscout$method_optim == 'optim'] = 'optim'
L = dscout$L_uniroot
L[!is.na(dscout$L_em)] = dscout$L_em[!is.na(dscout$L_em)]
L[!is.na(dscout$L_optim)] = dscout$L_optim[!is.na(dscout$L_optim)]
Time = dscout$Time_uniroot
Time[!is.na(dscout$Time_em)] = dscout$Time_em[!is.na(dscout$Time_em)]
Time[!is.na(dscout$Time_optim)] = dscout$Time_optim[!is.na(dscout$Time_optim)]
dscout = cbind(dscout, method, L, Time)
dscout = dscout[, -c(5:13)]

library(dplyr)

Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

library(knitr)
library(kableExtra)
dscout = readRDS('output/dsc_susie_z_v_output.rds')

We randomly generate X from N(0,1), n = 1200, p = 1000.

We randomly generate the response based on different number of signals (1, 3, 5, 10), pve (0.01, 0.2, 0.6, 0.8), whether the signals have the same effect size. We fit SuSiE model with L = 5 and 10. We perform simulations to compare three methods uniroot, em and optim. There are 100 replicates in the simulation. Therefore 19200 models in total.

All SuSiE models converge.

sum(dscout$converged)

[1] 19200

uniroot.obj = dscout$objective[dscout$method == 'uniroot']
em.obj = dscout$objective[dscout$method == 'em']
optim.obj = dscout$objective[dscout$method == 'optim']

hist(em.obj - uniroot.obj, main='Objective EM-uniroot', breaks=50)

hist(optim.obj - uniroot.obj, main='Objective optim-uniroot', breaks=50)

hist(optim.obj - em.obj, main='Objective optim-em', breaks=50)

em_uni = optim_uni = optim_em = matrix(NA,1,2)
weight = unique(dscout$effect_weight)
for(j in 1:2){
  tmp = dscout %>% filter(effect_weight == weight[j])
  uniroot.obj = tmp$objective[tmp$method == 'uniroot']
  em.obj = tmp$objective[tmp$method == 'em']
  optim.obj = tmp$objective[tmp$method == 'optim']
  em_uni[1,j] = sum(em.obj > uniroot.obj)/3200
  optim_uni[1,j] = sum(optim.obj > uniroot.obj)/3200
  optim_em[1,j] = sum(optim.obj > em.obj)/3200
}
colnames(em_uni) = colnames(optim_uni) = colnames(optim_em) = paste0('equal_', c('T', 'F'))

em_uni %>% kable() %>% kable_styling()

optim_uni %>% kable() %>% kable_styling()

optim_em %>% kable() %>% kable_styling()

library(ggplot2)
p <- ggplot(dscout, aes(x=method, y=Time)) + facet_wrap(~effect_weight)+ geom_violin(trim = FALSE) + coord_flip() + stat_summary(fun.y=mean, geom="point", shape=23, size=2)
p

sessionInfo()

R version 3.5.1 (2018-07-02)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS  10.14.3

Matrix products: default
BLAS: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] ggplot2_3.1.0    bindrcpp_0.2.2   kableExtra_1.0.1 knitr_1.20      
[5] dplyr_0.7.8     

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.0        plyr_1.8.4        highr_0.7        
 [4] compiler_3.5.1    pillar_1.3.1      git2r_0.24.0     
 [7] workflowr_1.1.1   bindr_0.1.1       R.methodsS3_1.7.1
[10] R.utils_2.7.0     tools_3.5.1       digest_0.6.18    
[13] gtable_0.2.0      evaluate_0.12     tibble_2.0.1     
[16] viridisLite_0.3.0 pkgconfig_2.0.2   rlang_0.3.1      
[19] rstudioapi_0.9.0  yaml_2.2.0        withr_2.1.2      
[22] stringr_1.3.1     httr_1.4.0        xml2_1.2.0       
[25] hms_0.4.2         grid_3.5.1        webshot_0.5.1    
[28] rprojroot_1.3-2   tidyselect_0.2.5  glue_1.3.0       
[31] R6_2.3.0          rmarkdown_1.11    purrr_0.2.5      
[34] readr_1.3.1       magrittr_1.5      whisker_0.3-2    
[37] backports_1.1.3   scales_1.0.0      htmltools_0.3.6  
[40] assertthat_0.2.0  rvest_0.3.2       colorspace_1.4-0 
[43] labeling_0.3      stringi_1.2.4     lazyeval_0.2.1   
[46] munsell_0.5.0     crayon_1.3.4      R.oo_1.22.0      

This reproducible R Markdown analysis was created with workflowr 1.1.1