Effects of Correction of Detection Rate

Last updated: 2018-07-16

workflowr checks: (Click a bullet for more information)

✔ 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(20180618)

The command set.seed(20180618) 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: 884f74e

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:    .Rhistory
    Ignored:    .Rproj.user/
    Ignored:    R/.Rhistory
    Ignored:    analysis/.Rhistory
    Ignored:    analysis/correcting_detection_rate/.Rhistory
    Ignored:    analysis/pipeline/.Rhistory

Untracked files:
    Untracked:  ..gif
    Untracked:  .DS_Store
    Untracked:  R/.DS_Store
    Untracked:  analysis/.DS_Store
    Untracked:  analysis/bibliography.bib
    Untracked:  analysis/clean_data/
    Untracked:  analysis/downsampling/
    Untracked:  analysis/dropouts/
    Untracked:  analysis/filteringtest.R
    Untracked:  analysis/old/
    Untracked:  analysis/pipeline/large_sets.pdf
    Untracked:  analysis/pipeline/temp_ari.txt
    Untracked:  analysis/pipeline/temp_time.txt
    Untracked:  analysis/sysdata_reference_test/
    Untracked:  analysis/tutorial_cache/
    Untracked:  analysis/writeup/cite.bib
    Untracked:  analysis/writeup/cite.log
    Untracked:  analysis/writeup/paper.aux
    Untracked:  analysis/writeup/paper.bbl
    Untracked:  analysis/writeup/paper.blg
    Untracked:  analysis/writeup/paper.log
    Untracked:  analysis/writeup/paper.out
    Untracked:  analysis/writeup/paper.synctex.gz
    Untracked:  analysis/writeup/paper.tex
    Untracked:  analysis/writeup/writeup.aux
    Untracked:  analysis/writeup/writeup.bbl
    Untracked:  analysis/writeup/writeup.blg
    Untracked:  analysis/writeup/writeup.dvi
    Untracked:  analysis/writeup/writeup.log
    Untracked:  analysis/writeup/writeup.out
    Untracked:  analysis/writeup/writeup.synctex.gz
    Untracked:  analysis/writeup/writeup.tex
    Untracked:  analysis/writeup/writeup2.aux
    Untracked:  analysis/writeup/writeup2.bbl
    Untracked:  analysis/writeup/writeup2.blg
    Untracked:  analysis/writeup/writeup2.log
    Untracked:  analysis/writeup/writeup2.out
    Untracked:  analysis/writeup/writeup2.pdf
    Untracked:  analysis/writeup/writeup2.synctex.gz
    Untracked:  analysis/writeup/writeup2.tex
    Untracked:  analysis/writeup/writeup3.aux
    Untracked:  analysis/writeup/writeup3.log
    Untracked:  analysis/writeup/writeup3.out
    Untracked:  analysis/writeup/writeup3.synctex.gz
    Untracked:  analysis/writeup/writeup3.tex
    Untracked:  data/unnecessary_in_building/

Unstaged changes:
    Modified:   R/SLSL.R
    Modified:   analysis/_site.yml
    Modified:   analysis/pipeline/large_sets.Rmd
    Modified:   analysis/pipeline/small_good_sets.Rmd
    Modified:   analysis/pipeline/small_good_sets_result.txt
    Modified:   analysis/pipeline/small_good_sets_time.txt
    Modified:   analysis/writeup/.DS_Store
    Modified:   src/SCNoisyClustering.cpp
    Modified:   src/SCNoisyClustering.o
    Modified:   src/SCNoisyClustering.so

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.

Expand here to see past versions:

File	Version	Author	Date	Message
Rmd	884f74e	tk382	2018-07-16	wflow_publish(“analysis/detection_rate_correction.Rmd”)
Rmd	66600ac	tk382	2018-07-16	add analysis files

Pollen

i = 2

load('../data/unnecessary_in_building/2_Pollen.RData')
X        = as.matrix(Pollen$x)
truth    = as.numeric(as.factor(Pollen$label))
numClust = length(unique(truth))
logX = log(X+1)

det = colSums(X!=0) / nrow(X)

det2 = qr(det)
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R))
tsne2 = Rtsne(t(logX))

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.7755788

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.8659285

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.7665684

Usoskin

i = 3

load('../data/unnecessary_in_building/3_Usoskin.RData')
X        = as.matrix(Usoskin$X)
truth    = as.numeric(as.factor(as.character(Usoskin$lab1)))
numClust = 4
rm(Usoskin)

logX = log(X+1)

det = colSums(X!=0) / nrow(X)

plot(irlba(logX,1)$v[,1]~log(det))

det2 = qr(cbind(rep(1, length(det)), log(det)))
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R), perplexity=20)
tsne2 = Rtsne(t(logX))

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.6188269

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.7345628

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.6348444

Buettner

i = 4

#read data
load('../data/unnecessary_in_building/4_Buettner.RData')
X        = as.matrix(Buettner$X)
truth    = as.numeric(as.factor(Buettner$label))
numClust = 3
rm(Buettner)

logX = log(X+1)
det = colSums(X!=0) / nrow(X)

det2 = qr(det)
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R), perplexity=20)
tsne2 = Rtsne(t(logX), perplexity=20)

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.4329975

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.428236

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.465002

Yan

i = 5

load('../data/unnecessary_in_building/5_Yan.rda')
X        = as.matrix(yan)
truth    = as.character(ann$cell_type1)
truth    = as.numeric(as.factor(truth))
numClust = 6
rm(ann, yan)

logX = log(X+1)

det = colSums(X!=0) / nrow(X)

det2 = qr(det)
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R), perplexity=20)
tsne2 = Rtsne(t(logX), perplexity=20)

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.9484199

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.9484199

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.7952752

Treutlein

i = 6

load('../data/unnecessary_in_building/6_Treutlein.rda')
X        = as.matrix(treutlein)
truth    = as.numeric(colnames(treutlein))
ind      = sort(truth, index.return=TRUE)$ix
X        = X[,ind]
truth    = truth[ind]
numClust = length(unique(truth))
rm(treutlein)
logX = log(X+1)

det = colSums(X!=0) / nrow(X)

det2 = qr(cbind(log(det), rep(1, length(det))))
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R), perplexity=10)
tsne2 = Rtsne(t(logX), perplexity=10)

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.794193

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.9603128

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.6517367

Chu (cell type)

i = 7

load('../data/unnecessary_in_building/7_Chu_celltype.Rdata')
X        = as.matrix(Chu_celltype$X)
truth    = as.numeric(as.factor(Chu_celltype$label))
numClust = 7
rm(Chu_celltype)

logX = log(X+1)

det = colSums(X!=0) / nrow(X)

det2 = qr(det)
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R))
tsne2 = Rtsne(t(logX))

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.7785884

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.7770962

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.775129

Chu (timecourse)

i = 8
load('../data/unnecessary_in_building/8_Chu_timecourse.Rdata')
X        = as.matrix(Chu_timecourse$X)
truth    = as.numeric(as.factor(Chu_timecourse$label))
numClust = length(unique(truth))

logX = log(X+1)

det = colSums(X!=0) / nrow(X)

det2 = qr(det)
R = t(qr.resid(det2, t(logX)))

pca1 = irlba(R,2); pca2 = irlba(logX,2)
dat = data.frame(pc1=c(pca1$v[,1], pca2$v[,1]), detection.rate=rep(det, 2), label=rep(c("After correction", "Before correction"), each=nrow(pca1$v)), true.label=as.factor(rep(truth,2)))
ggplot(dat, aes(x=pc1, y=detection.rate, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("PCA")

tsne1 = Rtsne(t(R))
tsne2 = Rtsne(t(logX))

dat = data.frame(v1 = c(tsne1$Y[,1], tsne2$Y[,1]), v2 = c(tsne1$Y[,2], tsne2$Y[,2]), label=rep(c("After correction", "Before correction"), each=nrow(tsne1$Y)), true.label = as.factor(rep(truth, 2)))
ggplot(dat, aes(x=v1, y=v2, col=true.label)) + facet_grid(~label) + geom_point() + ggtitle("tSNE")

set.seed(1); res1 = SLSL(R, log=F, filter=F)
adj.rand.index(res1$result, truth)

[1] 0.7321747

set.seed(1); res2 = SLSL(X, log=T, filter=F, correct_detection_rate = F)
adj.rand.index(res2$result, truth)

[1] 0.7276994

set.seed(1); res3 = SLSL(X, log=T, filter=F, correct_detection_rate = T)
adj.rand.index(res3$result, truth)

[1] 0.9692303

Session information

sessionInfo()

R version 3.5.1 (2018-07-02)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Sierra 10.12.5

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] stargazer_5.2.2         abind_1.4-5            
 [3] broom_0.4.5             diceR_0.5.1            
 [5] Rtsne_0.13              fossil_0.3.7           
 [7] shapefiles_0.7          foreign_0.8-70         
 [9] maps_3.3.0              sp_1.2-7               
[11] reshape_0.8.7           dplyr_0.7.6            
[13] ggplot2_3.0.0           irlba_2.3.2            
[15] Matrix_1.2-14           quadprog_1.5-5         
[17] inline_0.3.15           matrixStats_0.53.1     
[19] SCNoisyClustering_0.1.0

loaded via a namespace (and not attached):
 [1] tidyselect_0.2.4  reshape2_1.4.3    purrr_0.2.4      
 [4] lattice_0.20-35   colorspace_1.3-2  htmltools_0.3.6  
 [7] yaml_2.1.19       rlang_0.2.0       R.oo_1.22.0      
[10] pillar_1.2.2      glue_1.2.0        withr_2.1.2      
[13] R.utils_2.6.0     bindrcpp_0.2.2    plyr_1.8.4       
[16] bindr_0.1.1       stringr_1.3.0     munsell_0.4.3    
[19] gtable_0.2.0      workflowr_1.1.1   R.methodsS3_1.7.1
[22] codetools_0.2-15  psych_1.8.3.3     evaluate_0.10.1  
[25] labeling_0.3      knitr_1.20        parallel_3.5.1   
[28] Rcpp_0.12.16      scales_0.5.0      backports_1.1.2  
[31] mnormt_1.5-5      digest_0.6.15     stringi_1.1.7    
[34] grid_3.5.1        rprojroot_1.3-2   tools_3.5.1      
[37] magrittr_1.5      lazyeval_0.2.1    tibble_1.4.2     
[40] tidyr_0.8.0       whisker_0.3-2     pkgconfig_2.0.1  
[43] assertthat_0.2.0  rmarkdown_1.9     R6_2.2.2         
[46] mclust_5.4        nlme_3.1-137      git2r_0.21.0     
[49] compiler_3.5.1

This reproducible R Markdown analysis was created with workflowr 1.1.1