Last updated: 2019-05-01
Checks: 6 0
Knit directory: dsc-susie-z/
This reproducible R Markdown analysis was created with workflowr (version 1.3.0). The Checks tab describes the reproducibility checks that were applied when the results were created. The Past versions tab lists the development history.
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.
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.
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.
Great job! Recording the operating system, R version, and package versions is critical for reproducibility.
Nice! There were no cached chunks for this analysis, so you can be confident that you successfully produced the results during this run.
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: analysis/.DS_Store
Ignored: data/.DS_Store
Ignored: output/.DS_Store
Untracked files:
Untracked: data/random_data_31.rds
Untracked: data/random_data_31_sim_gaussian_35.rds
Untracked: data/random_data_31_sim_gaussian_35_get_sumstats_1.rds
Untracked: data/small_data_1.ld_in_file.in.ld
Untracked: data/small_data_1.ld_out_file.out.ld
Untracked: data/small_data_132.ld_in_file.in.ld
Untracked: data/small_data_132.ld_out_file.out.ld
Untracked: data/small_data_132_sim_gaussian_12.rds
Untracked: data/small_data_132_sim_gaussian_12_get_sumstats_1.rds
Untracked: data/small_data_1_sim_gaussian_2.rds
Untracked: data/small_data_1_sim_gaussian_2_get_sumstats_1.rds
Untracked: data/small_data_46.rds
Untracked: data/small_data_46_sim_gaussian_10.rds
Untracked: data/small_data_46_sim_gaussian_10_get_sumstats_2.rds
Untracked: data/small_data_69.ld_in_file.in.ld
Untracked: data/small_data_69.ld_out_file.out.ld
Untracked: data/small_data_69.rds
Untracked: data/small_data_69_sim_gaussian_3.rds
Untracked: data/small_data_69_sim_gaussian_3_get_sumstats_1.rds
Untracked: data/small_data_69_sim_gaussian_3_get_sumstats_1_susie_z_1.rds
Untracked: data/small_data_69_sim_gaussian_3_get_sumstats_1_susie_z_2.rds
Untracked: figure/
Untracked: output/dscoutProblem475.rds
Untracked: output/dscoutProblem75.rds
Untracked: output/finemap_compare_random_data_null_dscout.rds
Untracked: output/finemap_compare_random_data_signal_dscout.rds
Untracked: output/finemap_compare_small_data_signal_dscout.rds
Untracked: output/finemap_compare_small_data_signal_dscout_RE8.rds
Untracked: output/r_compare_FINEMAP_PIP_ROC.rds
Untracked: output/r_compare_add_z_FINEMAP_PIP_ROC.rds
Untracked: output/r_compare_add_z_FINEMAP_ROC.rds
Untracked: output/r_compare_add_z_SuSiEb_ROC.rds
Untracked: output/r_compare_add_z_SuSiErss_ROC.rds
Untracked: output/r_compare_add_z_dscout_susie_finemap_tibble.rds
Untracked: output/r_compare_add_z_lambda_FINEMAP_ROC.rds
Untracked: output/r_compare_add_z_lambda_SuSiEb_ROC.rds
Untracked: output/r_compare_add_z_lambda_SuSiErss_ROC.rds
Untracked: output/r_compare_add_z_lambda_dscout_susie_finemap_tibble.rds
Untracked: output/r_compare_dscout_susie_finemappip_tibble.rds
Untracked: output/r_compare_dscout_susie_finemappip_truth_tibble.rds
Untracked: output/r_compare_susieb_PIP_ROC.rds
Untracked: output/r_compare_susiepip_tibble.rds
Untracked: output/r_compare_susierss_PIP_ROC.rds
Untracked: output/random_data_100_sim_gaussian_null_1_get_sumstats_1_finemap_1.rds
Untracked: output/random_data_31_35_fit_em.rds
Untracked: output/random_data_76.rds
Untracked: output/random_data_76_sim_gaussian_8.rds
Untracked: output/random_data_76_sim_gaussian_8_get_sumstats_1.rds
Untracked: output/small_data_42_sim_gaussian_36_get_sumstats_2_susie_z_2.rds
Untracked: output/small_data_92_sim_gaussian_30_get_sumstats_2_susie_z_2.rds
Unstaged changes:
Modified: analysis/SuSiErssNotConverge.Rmd
Modified: analysis/SusieZPerformance.Rmd
Modified: analysis/SusieZPerformanceRE3.Rmd
Modified: output/dsc_susie_z_v_output.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.
These are the previous versions of the R Markdown and HTML files. If you’ve configured a remote Git repository (see ?wflow_git_remote), click on the hyperlinks in the table below to view them.
| File | Version | Author | Date | Message |
|---|---|---|---|---|
| html | f6bbfda | zouyuxin | 2019-04-30 | Build site. |
| Rmd | 1becb3f | zouyuxin | 2019-04-30 | wflow_publish(“analysis/r_compare_add_z_lambda_susie.Rmd”) |
The design matrix X are real human genotype data from GTEx project, the 150 data in dsc-finemap repo. We simulate under various number of causal variables (1,2) and total percentage of variance explained (0.1, 0.2). We set effect size of each causal variable to be equal. Using the summary statistics from univariate regression, we fit SuSiE model using in-sample/out-sample correlation matrix, and compare their results.
library(dscrutils)
library(tibble)Warning: package 'tibble' was built under R version 3.5.2library(kableExtra)
library(dplyr)
Attaching package: 'dplyr'The following objects are masked from 'package:stats':
filter, lagThe following objects are masked from 'package:base':
intersect, setdiff, setequal, uniondscout = dscquery('r_compare_add_z_lambda', targets = 'sim_gaussian sim_gaussian.pve sim_gaussian.n_signal sim_gaussian.meta data.N_in method.ld_method method.estimate_residual_variance method.lamb score_susie.total score_susie.valid score_susie.size score_susie.purity score_susie.top score_susie.converged score_susie.pip score_finemap.pip', omit.filenames = FALSE, groups = c("method: susie_bhat susie_bhat_add_z susie_rss susie_rss_add_z finemap finemap_add_z"))
dscout.tibble = as_tibble(dscout)dscout = readRDS('output/r_compare_add_z_lambda_dscout_susie_finemap_tibble.rds')
dscout$add_z = rep(FALSE, nrow(dscout))
dscout$add_z[dscout$method == 'finemap_add_z'] = TRUE
dscout$add_z[dscout$method == 'susie_bhat_add_z'] = TRUE
dscout$add_z[dscout$method == 'susie_rss_add_z'] = TRUE
dscout = dscout[,-6]
colnames(dscout) = c('DSC', 'filename','pve', 'n_signal', 'meta','N_in', 'method', 'ld_method', 'lambda', 'estimate_residual_variance', 'total', 'valid', 'size', 'purity', 'top', 'converged', 'susie.pip','finemap.pip','add_z')dscout.susierss = rbind(dscout[dscout$method == 'susie_rss',], dscout[dscout$method == 'susie_rss_add_z',])
dscout.susieb = rbind(dscout[dscout$method == 'susie_bhat',], dscout[dscout$method == 'susie_bhat_add_z',])
dscout.finemap = rbind(dscout[dscout$method == 'finemap',], dscout[dscout$method == 'finemap_add_z',])dscout.susieb.suc = dscout.susieb %>% filter(converged == 1)
dscout.susieb.in_sample = dscout.susieb.suc %>% filter(ld_method == 'in_sample') %>% filter(add_z == FALSE)
dscout.susieb.out_sample = dscout.susieb.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == FALSE)
dscout.susieb.out_sample.addz = dscout.susieb.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == TRUE)There are 3 models from susie_bhat fail to converge. There are 16 cases with out-sample R failed (out of 600). The estimated residual variance becomes negative.
converge.summary = aggregate(converged ~ ld_method + add_z + estimate_residual_variance, dscout.susieb, sum)
converge.summary$Fail = 600 - converge.summary$converged
Fail = converge.summary[converge.summary$Fail!=0,]
Fail[,-4] %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive"), full_width = F) | ld_method | add_z | estimate_residual_variance | Fail | |
|---|---|---|---|---|
| 2 | out_sample | FALSE | FALSE | 3 |
| 5 | out_sample | FALSE | TRUE | 15 |
| 6 | out_sample | TRUE | TRUE | 1 |
Estimate residual variance
purity.susieb.in_sample.5 = aggregate(purity~pve+n_signal, dscout.susieb.in_sample %>% filter(estimate_residual_variance==TRUE), mean)
colnames(purity.susieb.in_sample.5)[colnames(purity.susieb.in_sample.5) == 'purity'] <- 'in_sample'
purity.susieb.out_sample.5 = aggregate(purity~pve+n_signal, dscout.susieb.out_sample %>% filter(estimate_residual_variance==TRUE), mean)
colnames(purity.susieb.out_sample.5)[colnames(purity.susieb.out_sample.5) == 'purity'] <- 'out_sample'
purity.susieb.out_sample.addz.5 = aggregate(purity~pve+n_signal, dscout.susieb.out_sample.addz %>% filter(estimate_residual_variance==TRUE), mean)
colnames(purity.susieb.out_sample.addz.5)[colnames(purity.susieb.out_sample.addz.5) == 'purity'] <- 'out_sample.addz'
purity.susieb.5 = cbind(purity.susieb.in_sample.5, purity.susieb.out_sample.5, purity.susieb.out_sample.addz.5)
purity.susieb.5 = purity.susieb.5[,-c(4,5,7,8)]
round(purity.susieb.5, 3) %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive"), full_width = F) | pve | n_signal | in_sample | out_sample | out_sample.addz |
|---|---|---|---|---|
| 0.1 | 1 | 0.751 | 0.768 | 0.762 |
| 0.2 | 1 | 0.956 | 0.947 | 0.956 |
| 0.1 | 2 | 0.460 | 0.440 | 0.404 |
| 0.2 | 2 | 0.898 | 0.902 | 0.862 |
valid.in = aggregate(valid ~ pve+n_signal, dscout.susieb.in_sample %>% filter(estimate_residual_variance==TRUE), sum)
total.in = aggregate(DSC~ pve+n_signal, dscout.susieb.in_sample %>% filter(estimate_residual_variance==TRUE), length)
total.in$total_true = total.in$DSC * total.in$n_signal
power.susieb.in = merge(valid.in, total.in)
power.susieb.in$in_sample = round(power.susieb.in$valid/(power.susieb.in$total_true), 3)
power.susieb.in = power.susieb.in[,-c(3,4,5)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susieb.out_sample %>% filter(estimate_residual_variance==TRUE), sum)
total.out = aggregate(DSC~ pve+n_signal, dscout.susieb.out_sample%>% filter(estimate_residual_variance==TRUE), length)
total.out$total_true = total.out$DSC * total.out$n_signal
power.susieb.out = merge(valid.out, total.out)
power.susieb.out$out_sample = round(power.susieb.out$valid/(power.susieb.out$total_true), 3)
power.susieb.out = power.susieb.out[,-c(3,4,5)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susieb.out_sample.addz %>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz = aggregate(DSC~ pve+n_signal, dscout.susieb.out_sample.addz %>% filter(estimate_residual_variance==TRUE), length)
total.out.addz$total_true = total.out.addz$DSC * total.out.addz$n_signal
power.susieb.out.addz = merge(valid.out.addz, total.out.addz)
power.susieb.out.addz$out_sample.addz = round(power.susieb.out.addz$valid/(power.susieb.out.addz$total_true), 3)
power.susieb.out.addz = power.susieb.out.addz[,-c(3,4,5)]
power.susieb = cbind(power.susieb.in, power.susieb.out, power.susieb.out.addz)
power.susieb = power.susieb[,-c(4,5,7,8)]
power.susieb %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz |
|---|---|---|---|---|
| 0.1 | 1 | 0.833 | 0.865 | 0.859 |
| 0.1 | 2 | 0.267 | 0.235 | 0.207 |
| 0.2 | 1 | 0.973 | 0.958 | 0.967 |
| 0.2 | 2 | 0.733 | 0.656 | 0.630 |
valid.in = aggregate(valid ~ pve+n_signal, dscout.susieb.in_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.in = aggregate(total~ pve+n_signal, dscout.susieb.in_sample%>% filter(estimate_residual_variance==TRUE), sum)
fdr.in = merge(valid.in, total.in)
fdr.in$in_sample = round((fdr.in$total - fdr.in$valid)/fdr.in$total, 4)
fdr.in = fdr.in[,-c(3,4)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susieb.out_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.out = aggregate(total~ pve+n_signal, dscout.susieb.out_sample%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out = merge(valid.out, total.out)
fdr.out$out_sample = round((fdr.out$total - fdr.out$valid)/fdr.out$total, 4)
fdr.out = fdr.out[,-c(3,4)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susieb.out_sample.addz%>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz = aggregate(total~ pve+n_signal, dscout.susieb.out_sample.addz%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out.addz = merge(valid.out.addz, total.out.addz)
fdr.out.addz$out_sample.addz = round((fdr.out.addz$total - fdr.out.addz$valid)/fdr.out.addz$total, 4)
fdr.out.addz = fdr.out.addz[,-c(3,4)]
fdr.susieb = cbind(fdr.in, fdr.out, fdr.out.addz)
fdr.susieb = fdr.susieb[,-c(4,5,7,8)]
fdr.susieb %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz |
|---|---|---|---|---|
| 0.1 | 1 | 0.0234 | 0.2050 | 0.0725 |
| 0.1 | 2 | 0.1753 | 0.3429 | 0.2439 |
| 0.2 | 1 | 0.0331 | 0.4385 | 0.1761 |
| 0.2 | 2 | 0.1200 | 0.3754 | 0.2125 |
# only converged results
dscout.susierss.suc = dscout.susierss %>% filter(converged == 1)
dscout.susierss.in_sample = dscout.susierss.suc %>% filter(ld_method == 'in_sample') %>% filter(add_z == FALSE) %>% filter(lambda == 0)
dscout.susierss.out_sample = dscout.susierss.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == FALSE) %>% filter(lambda == 0)
dscout.susierss.out_sample.addz = dscout.susierss.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == TRUE) %>% filter(lambda == 0)
dscout.susierss.out_sample.lambda = dscout.susierss.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == FALSE) %>% filter(lambda == 1e-6)
dscout.susierss.out_sample.addz.lambda = dscout.susierss.suc %>% filter(ld_method == 'out_sample') %>% filter(add_z == TRUE) %>% filter(lambda == 1e-6)There are cases fail to converge in susie_rss.
converge.summary = aggregate(converged ~ ld_method + add_z + estimate_residual_variance + lambda, dscout.susierss, sum)
converge.summary$NotConverge = 600 - converge.summary$converged
NotConverge = converge.summary[converge.summary$NotConverge!=0,]
NotConverge[,-5] %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive"), full_width = F) | ld_method | add_z | estimate_residual_variance | lambda | NotConverge | |
|---|---|---|---|---|---|
| 2 | out_sample | FALSE | FALSE | 0e+00 | 2 |
| 5 | out_sample | FALSE | TRUE | 0e+00 | 2 |
| 8 | out_sample | FALSE | FALSE | 1e-06 | 1 |
| 11 | out_sample | FALSE | TRUE | 1e-06 | 1 |
Estimate residual variance
purity.susierss.in_sample = round(aggregate(purity~pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==TRUE), mean), 3)
colnames(purity.susierss.in_sample)[colnames(purity.susierss.in_sample) == 'purity'] <- 'in_sample'
purity.susierss.out_sample = round(aggregate(purity~pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==TRUE), mean), 3)
colnames(purity.susierss.out_sample)[colnames(purity.susierss.out_sample) == 'purity'] <- 'out_sample'
purity.susierss.out_sample.addz = round(aggregate(purity~pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==TRUE), mean), 3)
colnames(purity.susierss.out_sample.addz)[colnames(purity.susierss.out_sample.addz) == 'purity'] <- 'out_sample.addz'
purity.susierss.out_sample.lambda = round(aggregate(purity~pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==TRUE), mean), 3)
colnames(purity.susierss.out_sample.lambda)[colnames(purity.susierss.out_sample.lambda) == 'purity'] <- 'out_sample.lambda'
purity.susierss.out_sample.addz.lambda = round(aggregate(purity~pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==TRUE), mean), 3)
colnames(purity.susierss.out_sample.addz.lambda)[colnames(purity.susierss.out_sample.addz.lambda) == 'purity'] <- 'out_sample.addz.lambda'
purity.susierss = cbind(purity.susierss.in_sample, purity.susierss.out_sample, purity.susierss.out_sample.addz, purity.susierss.out_sample.lambda, purity.susierss.out_sample.addz.lambda)
purity.susierss = purity.susierss[,-c(4,5,7,8,10,11,13,14)]
purity.susierss %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive"), full_width = F) | pve | n_signal | in_sample | out_sample | out_sample.addz | out_sample.lambda | out_sample.addz.lambda |
|---|---|---|---|---|---|---|
| 0.1 | 1 | 0.758 | 0.724 | 0.771 | 0.719 | 0.772 |
| 0.2 | 1 | 0.961 | 0.937 | 0.958 | 0.936 | 0.958 |
| 0.1 | 2 | 0.465 | 0.382 | 0.412 | 0.386 | 0.406 |
| 0.2 | 2 | 0.892 | 0.829 | 0.841 | 0.828 | 0.830 |
Estimate residual variance
valid.in = aggregate(valid ~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.in = aggregate(DSC~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==TRUE), length)
total.in$total_true = total.in$DSC * total.in$n_signal
power.susierss.in = merge(valid.in, total.in)
power.susierss.in$in_sample = round(power.susierss.in$valid/(power.susierss.in$total_true), 3)
power.susierss.in = power.susierss.in[,-c(3,4,5)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.out = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==TRUE), length)
total.out$total_true = total.out$DSC * total.out$n_signal
power.susierss.out = merge(valid.out, total.out)
power.susierss.out$out_sample = round(power.susierss.out$valid/(power.susierss.out$total_true), 3)
power.susierss.out = power.susierss.out[,-c(3,4,5)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==TRUE), length)
total.out.addz$total_true = total.out.addz$DSC * total.out.addz$n_signal
power.susierss.out.addz = merge(valid.out.addz, total.out.addz)
power.susierss.out.addz$out_sample.addz = round(power.susierss.out.addz$valid/(power.susierss.out.addz$total_true), 3)
power.susierss.out.addz = power.susierss.out.addz[,-c(3,4,5)]
valid.out.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==TRUE), sum)
total.out.lambda = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==TRUE), length)
total.out.lambda$total_true = total.out.lambda$DSC * total.out.lambda$n_signal
power.susierss.out.lambda = merge(valid.out.lambda, total.out.lambda)
power.susierss.out.lambda$out_sample.lambda = round(power.susierss.out.lambda$valid/(power.susierss.out.lambda$total_true), 3)
power.susierss.out.lambda = power.susierss.out.lambda[,-c(3,4,5)]
valid.out.addz.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz.lambda = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==TRUE), length)
total.out.addz.lambda$total_true = total.out.addz.lambda$DSC * total.out.addz.lambda$n_signal
power.susierss.out.addz.lambda = merge(valid.out.addz.lambda, total.out.addz.lambda)
power.susierss.out.addz.lambda$out_sample.addz.lambda = round(power.susierss.out.addz.lambda$valid/(power.susierss.out.addz.lambda$total_true), 3)
power.susierss.out.addz.lambda = power.susierss.out.addz.lambda[,-c(3,4,5)]
power.susierss = cbind(power.susierss.in, power.susierss.out, power.susierss.out.addz, power.susierss.out.lambda, power.susierss.out.addz.lambda)
power.susierss = power.susierss[,-c(4,5,7,8,10,11,13,14)]
power.susierss %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz | out_sample.lambda | out_sample.addz.lambda |
|---|---|---|---|---|---|---|
| 0.1 | 1 | 0.827 | 0.753 | 0.860 | 0.747 | 0.860 |
| 0.1 | 2 | 0.263 | 0.201 | 0.210 | 0.203 | 0.210 |
| 0.2 | 1 | 0.953 | 0.813 | 0.953 | 0.813 | 0.953 |
| 0.2 | 2 | 0.733 | 0.574 | 0.607 | 0.577 | 0.600 |
NOT estimate residual variance
valid.in = aggregate(valid ~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==FALSE), sum)
total.in = aggregate(DSC~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==FALSE), length)
total.in$total_true = total.in$DSC * total.in$n_signal
power.susierss.in = merge(valid.in, total.in)
power.susierss.in$in_sample = round(power.susierss.in$valid/(power.susierss.in$total_true), 3)
power.susierss.in = power.susierss.in[,-c(3,4,5)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==FALSE), sum)
total.out = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==FALSE), length)
total.out$total_true = total.out$DSC * total.out$n_signal
power.susierss.out = merge(valid.out, total.out)
power.susierss.out$out_sample = round(power.susierss.out$valid/(power.susierss.out$total_true), 3)
power.susierss.out = power.susierss.out[,-c(3,4,5)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==FALSE), sum)
total.out.addz = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==FALSE), length)
total.out.addz$total_true = total.out.addz$DSC * total.out.addz$n_signal
power.susierss.out.addz = merge(valid.out.addz, total.out.addz)
power.susierss.out.addz$out_sample.addz = round(power.susierss.out.addz$valid/(power.susierss.out.addz$total_true), 3)
power.susierss.out.addz = power.susierss.out.addz[,-c(3,4,5)]
valid.out.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==FALSE), sum)
total.out.lambda = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==FALSE), length)
total.out.lambda$total_true = total.out.lambda$DSC * total.out.lambda$n_signal
power.susierss.out.lambda = merge(valid.out.lambda, total.out.lambda)
power.susierss.out.lambda$out_sample.lambda = round(power.susierss.out.lambda$valid/(power.susierss.out.lambda$total_true), 3)
power.susierss.out.lambda = power.susierss.out.lambda[,-c(3,4,5)]
valid.out.addz.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==FALSE), sum)
total.out.addz.lambda = aggregate(DSC~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==FALSE), length)
total.out.addz.lambda$total_true = total.out.addz.lambda$DSC * total.out.addz.lambda$n_signal
power.susierss.out.addz.lambda = merge(valid.out.addz.lambda, total.out.addz.lambda)
power.susierss.out.addz.lambda$out_sample.addz.lambda = round(power.susierss.out.addz.lambda$valid/(power.susierss.out.addz.lambda$total_true), 3)
power.susierss.out.addz.lambda = power.susierss.out.addz.lambda[,-c(3,4,5)]
power.susierss = cbind(power.susierss.in, power.susierss.out, power.susierss.out.addz, power.susierss.out.lambda, power.susierss.out.addz.lambda)
power.susierss = power.susierss[,-c(4,5,7,8,10,11,13,14)]
power.susierss %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz | out_sample.lambda | out_sample.addz.lambda |
|---|---|---|---|---|---|---|
| 0.1 | 1 | 0.820 | 0.753 | 0.860 | 0.747 | 0.860 |
| 0.1 | 2 | 0.243 | 0.201 | 0.200 | 0.203 | 0.200 |
| 0.2 | 1 | 0.953 | 0.813 | 0.953 | 0.813 | 0.953 |
| 0.2 | 2 | 0.690 | 0.574 | 0.563 | 0.577 | 0.563 |
Estimate residual variance
valid.in = aggregate(valid ~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.in = aggregate(total~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==TRUE), sum)
fdr.in = merge(valid.in, total.in)
fdr.in$in_sample = round((fdr.in$total - fdr.in$valid)/fdr.in$total, 4)
fdr.in = fdr.in[,-c(3,4)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==TRUE), sum)
total.out = aggregate(total~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out = merge(valid.out, total.out)
fdr.out$out_sample = round((fdr.out$total - fdr.out$valid)/fdr.out$total, 4)
fdr.out = fdr.out[,-c(3,4)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out.addz = merge(valid.out.addz, total.out.addz)
fdr.out.addz$out_sample.addz = round((fdr.out.addz$total - fdr.out.addz$valid)/fdr.out.addz$total, 4)
fdr.out.addz = fdr.out.addz[,-c(3,4)]
valid.out.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==TRUE), sum)
total.out.lambda = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out.lambda = merge(valid.out.lambda, total.out.lambda)
fdr.out.lambda$out_sample.lambda = round((fdr.out.lambda$total - fdr.out.lambda$valid)/fdr.out.lambda$total, 4)
fdr.out.lambda = fdr.out.lambda[,-c(3,4)]
valid.out.addz.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==TRUE), sum)
total.out.addz.lambda = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==TRUE), sum)
fdr.out.addz.lambda = merge(valid.out.addz.lambda, total.out.addz.lambda)
fdr.out.addz.lambda$out_sample.addz = round((fdr.out.addz.lambda$total - fdr.out.addz.lambda$valid)/fdr.out.addz.lambda$total, 4)
fdr.out.addz.lambda = fdr.out.addz.lambda[,-c(3,4)]
fdr.susierss = cbind(fdr.in, fdr.out, fdr.out.addz, fdr.out.lambda, fdr.out.addz.lambda)
fdr.susierss = fdr.susierss[,-c(4,5,7,8,10,11,13,14)]
fdr.susierss %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz | out_sample.lambda | out_sample.addz.1 |
|---|---|---|---|---|---|---|
| 0.1 | 1 | 0.0388 | 0.1504 | 0.0915 | 0.1515 | 0.0915 |
| 0.1 | 2 | 0.1856 | 0.2593 | 0.2317 | 0.2651 | 0.2410 |
| 0.2 | 1 | 0.0467 | 0.2561 | 0.1538 | 0.2469 | 0.1538 |
| 0.2 | 2 | 0.1200 | 0.2366 | 0.1947 | 0.2321 | 0.2000 |
NOT estimate residual variance
valid.in = aggregate(valid ~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==FALSE), sum)
total.in = aggregate(total~ pve+n_signal, dscout.susierss.in_sample%>% filter(estimate_residual_variance==FALSE), sum)
fdr.in = merge(valid.in, total.in)
fdr.in$in_sample = round((fdr.in$total - fdr.in$valid)/fdr.in$total, 4)
fdr.in = fdr.in[,-c(3,4)]
valid.out = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==FALSE), sum)
total.out = aggregate(total~ pve+n_signal, dscout.susierss.out_sample%>% filter(estimate_residual_variance==FALSE), sum)
fdr.out = merge(valid.out, total.out)
fdr.out$out_sample = round((fdr.out$total - fdr.out$valid)/fdr.out$total, 4)
fdr.out = fdr.out[,-c(3,4)]
valid.out.addz = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==FALSE), sum)
total.out.addz = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.addz%>% filter(estimate_residual_variance==FALSE), sum)
fdr.out.addz = merge(valid.out.addz, total.out.addz)
fdr.out.addz$out_sample.addz = round((fdr.out.addz$total - fdr.out.addz$valid)/fdr.out.addz$total, 4)
fdr.out.addz = fdr.out.addz[,-c(3,4)]
valid.out.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==FALSE), sum)
total.out.lambda = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.lambda%>% filter(estimate_residual_variance==FALSE), sum)
fdr.out.lambda = merge(valid.out.lambda, total.out.lambda)
fdr.out.lambda$out_sample.lambda = round((fdr.out.lambda$total - fdr.out.lambda$valid)/fdr.out.lambda$total, 4)
fdr.out.lambda = fdr.out.lambda[,-c(3,4)]
valid.out.addz.lambda = aggregate(valid ~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==FALSE), sum)
total.out.addz.lambda = aggregate(total~ pve+n_signal, dscout.susierss.out_sample.addz.lambda%>% filter(estimate_residual_variance==FALSE), sum)
fdr.out.addz.lambda = merge(valid.out.addz.lambda, total.out.addz.lambda)
fdr.out.addz.lambda$out_sample.addz = round((fdr.out.addz.lambda$total - fdr.out.addz.lambda$valid)/fdr.out.addz.lambda$total, 4)
fdr.out.addz.lambda = fdr.out.addz.lambda[,-c(3,4)]
fdr.susierss = cbind(fdr.in, fdr.out, fdr.out.addz, fdr.out.lambda, fdr.out.addz.lambda)
fdr.susierss = fdr.susierss[,-c(4,5,7,8,10,11,13,14)]
fdr.susierss %>% kable() %>% kable_styling(bootstrap_options = c("striped", "condensed", "responsive","bordered"), full_width = F)| pve | n_signal | in_sample | out_sample | out_sample.addz | out_sample.lambda | out_sample.addz.1 |
|---|---|---|---|---|---|---|
| 0.1 | 1 | 0.0315 | 0.1504 | 0.0444 | 0.1515 | 0.0444 |
| 0.1 | 2 | 0.1889 | 0.2593 | 0.2405 | 0.2651 | 0.2500 |
| 0.2 | 1 | 0.0467 | 0.2561 | 0.1062 | 0.2469 | 0.1062 |
| 0.2 | 2 | 0.1229 | 0.2366 | 0.1716 | 0.2321 | 0.1756 |
sessionInfo()R version 3.5.1 (2018-07-02)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS 10.14.4
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] bindrcpp_0.2.2 dplyr_0.7.8 kableExtra_1.0.1 tibble_2.0.1
[5] dscrutils_0.3.8
loaded via a namespace (and not attached):
[1] Rcpp_1.0.0 highr_0.7 pillar_1.3.1
[4] compiler_3.5.1 git2r_0.24.0 workflowr_1.3.0
[7] bindr_0.1.1 tools_3.5.1 digest_0.6.18
[10] evaluate_0.12 viridisLite_0.3.0 pkgconfig_2.0.2
[13] rlang_0.3.1 rstudioapi_0.9.0 yaml_2.2.0
[16] stringr_1.3.1 httr_1.4.0 knitr_1.20
[19] xml2_1.2.0 fs_1.2.6 hms_0.4.2
[22] tidyselect_0.2.5 rprojroot_1.3-2 webshot_0.5.1
[25] glue_1.3.0 R6_2.3.0 rmarkdown_1.11
[28] purrr_0.2.5 readr_1.3.1 magrittr_1.5
[31] whisker_0.3-2 backports_1.1.3 scales_1.0.0
[34] htmltools_0.3.6 assertthat_0.2.0 rvest_0.3.2
[37] colorspace_1.4-0 stringi_1.2.4 munsell_0.5.0
[40] crayon_1.3.4