Kevin Blighe 2019-01-29
CorLevelPlot provides a quick and colourful way to visualise statistically significant correlations between any combination of categorical and continuous variables. Moreover, the statistical significancies of these correlations are also provided.
Example CorLevelPlot plotsInstall and load CorLevelPlot:
devtools::install_github("kevinblighe/CorLevelPlot") Example 1: WGCNA (weighted gene co-expression network analysis) simulated data:
The following code taken from Tutorial for the WGCNA package for R - 1. Simulation of expression and trait data
# simulate trait-to-eigengene data
no.obs <- 50
ESturquoise <- 0; ESbrown <- -0.6; ESgreen <- 0.6; ESyellow <- 0
ESvector <- c(ESturquoise, ESbrown, ESgreen, ESyellow)
nGenes1 <- 3000
simulateProportions1 <- c(0.2, 0.15, 0.08, 0.06, 0.04)
set.seed(1)
MEgreen <- rnorm(no.obs)
scaledy <- MEgreen * ESgreen + sqrt(1 - ESgreen ^ 2) * rnorm(no.obs)
y <- ifelse( scaledy > median(scaledy), 2, 1)
MEturquoise <- ESturquoise * scaledy + sqrt(1 - ESturquoise ^ 2) * rnorm(no.obs)
MEblue <- 0.6 * MEturquoise + sqrt(1 - 0.6 ^ 2) * rnorm(no.obs)
MEbrown <- ESbrown * scaledy + sqrt(1 - ESbrown ^ 2) * rnorm(no.obs)
MEyellow <- ESyellow * scaledy + sqrt(1 - ESyellow ^ 2) * rnorm(no.obs)
ModuleEigengeneNetwork1 <- data.frame(y, MEturquoise, MEblue, MEbrown, MEgreen, MEyellow)
CorLevelPlot(data = ModuleEigengeneNetwork1,
x = c("y", "MEturquoise", "MEblue", "MEbrown", "MEgreen", "MEyellow"),
y = c("y", "MEturquoise", "MEblue", "MEbrown", "MEgreen", "MEyellow"),
titleX = "X correlates",
cexTitleX = 3.0,
rotTitleX = 0,
colTitleX = "forestgreen",
fontTitleX = 2,
titleY = "Y\ncorrelates",
cexTitleY = 4.0,
rotTitleY = 100,
colTitleY = "gold",
fontTitleY = 4,
cexLabX = 1.0,
rotLabX = 45,
colLabX = "grey20",
fontLabX = 1,
cexLabY = 1.0,
rotLabY = 30,
colLabY = "royalblue",
fontLabY = 1,
posLab = "bottomleft",
col = c("blue4", "blue3", "blue2", "blue1", "white", "red1", "red2", "red3", "red4"),
posColKey = "right",
cexLabColKey = 1.2,
cexCorval = 1.0,
fontCorval = 4,
main = "WGCNA example",
scale = FALSE,
cexMain = 2,
rotMain = 360,
colMain = "red4",
fontMain = 4,
corFUN = "pearson",
corUSE = "pairwise.complete.obs",
signifSymbols = c("***", "**", "*", ""),
signifCutpoints = c(0, 0.001, 0.01, 0.05, 1),
colFrame = "white",
plotRsquared = FALSE)
Example 2: Iris dataset principal components analysis:
library(datasets)
data(iris)
# order the categories in the 'Species' column
# CorLevelPlot will conver these to 1, 2, 3, ...
iris$Species <- as.numeric(factor(iris$Species, levels=c("setosa", "versicolor", "virginica")))
i <- CorLevelPlot(data = iris,
x = c("Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width", "Species"),
y = c("Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width", "Species"),
col = c("white", "cornsilk1", "yellow", "gold", "forestgreen", "darkgreen"),
cexCorval = 1.2,
fontCorval = 2,
posLab = "all",
rotLabX = 45,
scale = TRUE,
main = bquote(Iris~r^2~correlates),
plotRsquared = TRUE)
pca <- stats::prcomp(iris[,c("Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width")])
df <- data.frame(pca$x, iris)
ii <- CorLevelPlot(data = df,
x = c("PC1", "PC2", "PC3", "PC4"),
y = c("Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width", "Species"),
cexTitleX = 2.0,
rotTitleX = 0,
fontTitleX = 2,
titleY = "Iris components",
cexTitleY = 2.0,
rotTitleY = 90,
fontTitleY = 2,
posLab = "topright",
col = c("blue1", "skyblue", "white", "pink", "red1"),
posColKey = "bottom",
cexLabColKey = 1.5,
cexCorval = 1.5,
fontCorval = 2,
rotLabX = 45,
scale = TRUE,
main = "Iris PC correlates",
colFrame = "white",
plotRsquared = FALSE)
require(rasterVis)
require(gridExtra)
require(grid)
grid.arrange(
arrangeGrob(i,
top = textGrob("A",
x = unit(0.05,"npc"),
y = unit(0.9,"npc"),
just = c("left","top"),
gp = gpar(fontsize=32))),
arrangeGrob(ii,
top = textGrob("B",
x = unit(0.05,"npc"),
y = unit(0.9,"npc"),
just = c("left","top"),
gp = gpar(fontsize=32))),
ncol = 2)
Example 3: World Health Organization (WHO) MONICA data:
library(DAAG)
data(monica)
# order the categorical variables
monica$outcome <- as.numeric(factor(monica$outcome, levels=c("dead", "live")))
monica$diabetes[monica$diabetes=="nk"] <- NA
monica$diabetes <- as.numeric(factor(monica$diabetes, levels=c("n", "y")))
monica$hichol[monica$hichol=="nk"] <- NA
monica$hichol <- as.numeric(factor(monica$hichol, levels=c("n", "y")))
monica$stroke[monica$stroke=="nk"] <- NA
monica$stroke <- as.numeric(factor(monica$stroke, levels=c("n", "y")))
monica$sex <- as.numeric(factor(monica$sex, levels=c("m", "f")))
monica$yronset <- as.numeric(factor(monica$yronset, levels=c("85","86","87","88","89","90","91","92","93")))
monica$highbp[monica$highbp=="nk"] <- NA
monica$highbp <- as.numeric(factor(monica$highbp, levels=c("n", "y")))
monica$angina[monica$angina=="nk"] <- NA
monica$angina <- as.numeric(factor(monica$angina, levels=c("n", "y")))
monica$hosp <- as.numeric(factor(monica$hosp, levels=c("n", "y")))
CorLevelPlot(data = monica,
x = c("outcome", "diabetes", "highbp", "hichol", "angina", "hosp"),
y = c("sex", "age", "yronset"),
col = c("darkblue", "blue2", "black", "red2", "darkred"),
cexCorval = 1.5,
colCorval = "white",
fontCorval = 2,
posLab = "bottomleft",
rotLabX = 45,
posColKey = "top",
cexLabColKey = 1.2,
scale = TRUE,
main = "World Health Organization",
colFrame = "white",
plotRsquared = FALSE)
## R version 3.5.2 (2018-12-20)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Ubuntu 16.04.5 LTS
##
## Matrix products: default
## BLAS: /usr/lib/atlas-base/atlas/libblas.so.3.0
## LAPACK: /usr/lib/atlas-base/atlas/liblapack.so.3.0
##
## locale:
## [1] LC_CTYPE=pt_BR.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_GB.UTF-8 LC_COLLATE=pt_BR.UTF-8
## [5] LC_MONETARY=en_GB.UTF-8 LC_MESSAGES=pt_BR.UTF-8
## [7] LC_PAPER=en_GB.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] grid stats graphics grDevices utils datasets methods
## [8] base
##
## other attached packages:
## [1] DAAG_1.22 gridExtra_2.3 rasterVis_0.45
## [4] latticeExtra_0.6-28 RColorBrewer_1.1-2 lattice_0.20-38
## [7] raster_2.8-4 sp_1.3-1 BiocInstaller_1.32.1
## [10] CorLevelPlot_0.99.0 knitr_1.21
##
## loaded via a namespace (and not attached):
## [1] Rcpp_1.0.0 compiler_3.5.2 highr_0.7
## [4] prettyunits_1.0.2 remotes_2.0.2 tools_3.5.2
## [7] digest_0.6.18 pkgbuild_1.0.2 pkgload_1.0.2
## [10] gtable_0.2.0 viridisLite_0.3.0 evaluate_0.12
## [13] memoise_1.1.0 rlang_0.3.1 cli_1.0.1
## [16] parallel_3.5.2 curl_3.3 yaml_2.2.0
## [19] hexbin_1.27.2 xfun_0.4 withr_2.1.2
## [22] stringr_1.3.1 desc_1.2.0 fs_1.2.6
## [25] devtools_2.0.1 rprojroot_1.3-2 glue_1.3.0
## [28] R6_2.3.0 processx_3.2.1 rmarkdown_1.11
## [31] sessioninfo_1.1.1 callr_3.1.1 magrittr_1.5
## [34] codetools_0.2-16 backports_1.1.3 ps_1.3.0
## [37] htmltools_0.3.6 usethis_1.4.0 assertthat_0.2.0
## [40] stringi_1.2.4 crayon_1.3.4 zoo_1.8-4
(Blighe 2018)
Blighe, Kevin. 2018. “CorLevelPlot: Visualise correlation results and test significancies of these.” https://github.com/kevinblighe.
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4