Branches of mechanical engineering: Add Together P-Values As Well As Significance Levels To Ggplots

(This article was commencement published on Easy Guides, too kindly contributed to R-bloggers)

In this article, we’ll clit how to easily i) compare means of ii or multiple groups; ii) too to automatically add p-values and significance levels to a ggplot (such every bit box plots, point plots, bar plots too occupation plots …).

Contents:

• Prerequisites
• Methods for comparing means
• R functions to add together p-values
• Compare ii independent groups
• Compare ii paired samples
• Compare to a greater extent than than ii groups
• Multiple grouping variables
• Other plot types
• Infos

Prerequisites

Install too charge required R packages

Required R package: ggpubr (version >= 0.1.3), for ggplot2-based publication attain plots.

• Install from CRAN as follow:

install.packages("ggpubr")

• Or, install the latest developmental version from GitHub as follow:

if(!require(devtools)) install.packages("devtools") devtools::install_github("kassambara/ggpubr")

• Load ggpubr:

library(ggpubr)

Official documentation of ggpubr is available at: http://www.sthda.com/english/rpkgs/ggpubr

Demo information sets

Data: ToothGrowth data sets.

data("ToothGrowth") head(ToothGrowth)

   len supp dose 1  4.2   VC  0.5 2 11.5   VC  0.5 three  7.3   VC  0.5 iv  5.8   VC  0.5 five  6.4   VC  0.5 half dozen 10.0   VC  0.5

Methods for comparing means

The measure methods to compare the agency of ii or to a greater extent than groups inward R, convey been largely described at: comparing agency inward R.
The close mutual methods for comparing agency include:

METHODS	R FUNCTION	DESCRIPTION
T-test	t.test()	Compare ii groups (parametric)
Wilcoxon test	wilcox.test()	Compare ii groups (non-parametric)
ANOVA	aov() or anova()	Compare multiple groups (parametric)
Kruskal-Wallis	kruskal.test()	Compare multiple groups (non-parametric)

Influenza A virus subtype H5N1 practical direct to compute too translate the results of each of these methods are provided at the next links:

• Comparing one-sample hateful to a measure known mean:
  
  • One-Sample T-test (parametric)
  • One-Sample Wilcoxon Test (non-parametric)
• Comparing the agency of ii independent groups:
  
  • Unpaired Two Samples T-test (parametric)
  • Unpaired Two-Samples Wilcoxon Test (non-parametric)
• Comparing the agency of paired samples:
  
  • Paired Samples T-test (parametric)
  • Paired Samples Wilcoxon Test (non-parametric)
• Comparing the agency of to a greater extent than than ii groups
  
  • Analysis of variance (ANOVA, parametric):
    
    • One-Way ANOVA Test inward R
    • Two-Way ANOVA Test inward R
  • Kruskal-Wallis Test inward R (non parametric alternative to one-way ANOVA)

R functions to add together p-values

Here nosotros introduce ii novel R functions inward the ggpubr package:

• compare_means(): slow to occupation solution to performs ane too multiple hateful comparisons.
• stat_compare_means(): slow to occupation solution to automatically add together p-values too significance levels to a ggplot.

compare_means()

As we’ll demo inward the adjacent sections, it has multiple useful options compared to the measure R functions.
The simplified format is every bit follow:

compare_means(formula, data, method = "wilcox.test", paired = FALSE,   group.by = NULL, ref.group = NULL, ...)

• formula: a formula of the form x group, where x is a numeric variable too grouping is a constituent amongst ane or multiple levels. For example, formula = TP53 cancer_group. It’s also possible to perform the attempt out for multiple response variables at the same time. For example, formula = c(TP53, PTEN) cancer_group.
• data: a data.frame containing the variables inward the formula.
• method: the type of test. Default is “wilcox.test”. Allowed values include:
  
  • “t.test” (parametric) and “wilcox.test”” (non-parametric). Perform comparing betwixt ii groups of samples. If the grouping variable contains to a greater extent than than ii levels, too then a pairwise comparing is performed.
  • “anova” (parametric) and “kruskal.test” (non-parametric). Perform one-way ANOVA attempt out comparing multiple groups.
• paired: a logical indicating whether yous desire a paired test. Used alone in t.test and in wilcox.test.
• group.by: variables used to grouping the information laid upward earlier applying the test. When specified the hateful comparisons volition move performed inward each subset of the information formed past times the dissimilar levels of the group.by variables.
• ref.group: a grapheme string specifying the reference group. If specified, for a given grouping variable, each of the grouping levels volition move compared to the reference grouping (i.e. control group). ref.group tin displace move also “.all.”. In this case, each of the grouping variable levels is compared to all (i.e. base-mean).

stat_compare_means()

This business office extends ggplot2 for adding hateful comparing p-values to a ggplot, such every bit box blots, point plots, bar plots too occupation plots.
The simplified format is every bit follow:

stat_compare_means(mapping = NULL, comparisons = NULL hide.ns = FALSE,                    label = NULL,  label.x = NULL, label.y = NULL,  ...)

• mapping: Set of aesthetic mappings created past times aes().
• comparisons: Influenza A virus subtype H5N1 listing of length-2 vectors. The entries inward the vector are either the names of 2 values on the x-axis or the 2 integers that jibe to the index of the groups of interest, to move compared.
• hide.ns: logical value. If TRUE, cover ns symbol when displaying significance levels.
• label: grapheme string specifying label type. Allowed values include “p.signif” (shows the significance levels), “p.format” (shows the formatted p value).
• label.x,label.y: numeric values. coordinates (in information units) to move used for absolute positioning of the label. If likewise brusk they volition move recycled.
• …: other arguments passed to the function compare_means() such as method, paired, ref.group.

Compare ii independent groups

Perform the test:

compare_means(len   supp, information = ToothGrowth)

# Influenza A virus subtype H5N1 tibble: 1 x 8     .y. group1 group2          p      p.adj p.format p.signif   method                                1   len     OJ     VC 0.06449067 0.06449067    0.064       ns Wilcoxon

By default method = “wilcox.test” (non-parametric test). You tin displace also specify method = “t.test” for a parametric t-test.

Returned value is a information frame amongst the next columns:

• .y.: the y variable used inward the test.
• p: the p-value
• p.adj: the adjusted p-value. Default value for p.adjust.method = “holm”
• p.format: the formatted p-value
• p.signif: the significance level.
• method: the statistical attempt out used to compare groups.

Create a box plot amongst p-values:

p <- ggboxplot(ToothGrowth, x = "supp", y = "len",           color = "supp", palette = "jco",           add together = "jitter") #  Add p-value p + stat_compare_means()  # Change method p + stat_compare_means(method = "t.test")

Add p-values too significance levels to ggplots

Note that, the p-value label seat tin displace move adjusted using the arguments: label.x, label.y, hjust too vjust.
The default p-value label displayed is obtained past times concatenating the method and the p columns of the returned information frame past times the function compare_means(). You tin displace specify other combinations using the aes() function.
For example,

• aes(label = ..p.format..) or aes(label = paste0(“p =”, ..p.format..)): display alone the formatted p-value (without the method name)
• aes(label = ..p.signif..): display alone the significance level.
• aes(label = paste0(..method.., “\n”, “p =”, ..p.format..)): Use occupation intermission (“\n”) betwixt the method cite too the p-value.

As an illustration, type this:

p + stat_compare_means( aes(label = ..p.signif..),                          label.x = 1.5, label.y = 40)

Add p-values too significance levels to ggplots

If yous prefer, it’s also possible to specify the argument label as a grapheme vector:

p + stat_compare_means( label = "p.signif", label.x = 1.5, label.y = 40)

Compare ii paired samples

Perform the test:

compare_means(len   supp, information = ToothGrowth, paired = TRUE)

# Influenza A virus subtype H5N1 tibble: 1 x 8     .y. group1 group2           p       p.adj p.format p.signif   method                                  1   len     OJ     VC 0.004312554 0.004312554   0.0043       ** Wilcoxon

Visualize paired information using the ggpaired() function:

ggpaired(ToothGrowth, x = "supp", y = "len",          color = "supp", line.color = "gray", line.size = 0.4,          palette = "jco")+   stat_compare_means(paired = TRUE)

Add p-values too significance levels to ggplots

Compare to a greater extent than than ii groups

• Global test:

# Global attempt out compare_means(len   dose,  information = ToothGrowth, method = "anova")

# Influenza A virus subtype H5N1 tibble: 1 x half dozen     .y.            p        p.adj p.format p.signif method                              1   len 9.532727e-16 9.532727e-16  9.5e-16     ****  Anova

Plot amongst global p-value:

# Default method = "kruskal.test" for multiple groups ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+   stat_compare_means()  # Change method to anova ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+   stat_compare_means(method = "anova")

Add p-values too significance levels to ggplots

• Pairwise comparisons. If the grouping variable contains to a greater extent than than ii levels, too then pairwise tests volition move performed automatically. The default method is “wilcox.test”. You tin displace modify this to “t.test”.

# Perorm pairwise comparisons compare_means(len   dose,  information = ToothGrowth)

# Influenza A virus subtype H5N1 tibble: three x 8     .y. group1 group2            p        p.adj p.format p.signif   method                                    1   len    0.5      1 7.020855e-06 1.404171e-05  7.0e-06     **** Wilcoxon 2   len    0.5      2 8.406447e-08 2.521934e-07  8.4e-08     **** Wilcoxon three   len      1      2 1.772382e-04 1.772382e-04  0.00018      *** Wilcoxon

# Visualize: Specify the comparisons yous desire my_comparisons <- list( c("0.5", "1"), c("1", "2"), c("0.5", "2") ) ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+    stat_compare_means(comparisons = my_comparisons)+ # Add pairwise comparisons p-value   stat_compare_means(label.y = 50)     # Add global p-value

Add p-values too significance levels to ggplots

If yous desire to specify the precise y location of bars, occupation the argument label.y:

ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+    stat_compare_means(comparisons = my_comparisons, label.y = c(29, 35, 40))+   stat_compare_means(label.y = 45)

Add p-values too significance levels to ggplots

(Adding bars, connecting compared groups, has been facilitated past times the ggsignifR bundle )

• Multiple pairwise tests against a reference group:

# Pairwise comparing against reference compare_means(len   dose,  information = ToothGrowth, ref.group = "0.5",               method = "t.test")

# Influenza A virus subtype H5N1 tibble: 2 x 8     .y. group1 group2            p        p.adj p.format p.signif method                                  1   len    0.5      1 6.697250e-09 6.697250e-09  6.7e-09     **** T-test 2   len    0.5      2 1.469534e-16 2.939068e-16  < 2e-16     **** T-test

# Visualize ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+   stat_compare_means(method = "anova", label.y = 40)+      # Add global p-value   stat_compare_means(label = "p.signif", method = "t.test",                      ref.group = "0.5")                    # Pairwise comparing against reference

Add p-values too significance levels to ggplots

• Multiple pairwise tests against all (base-mean):

# Comparison of each grouping against base-mean compare_means(len   dose,  information = ToothGrowth, ref.group = ".all.",               method = "t.test")

# Influenza A virus subtype H5N1 tibble: three x 8     .y. group1 group2            p        p.adj p.format p.signif method                                  1   len  .all.    0.5 1.244626e-06 3.733877e-06  1.2e-06     **** T-test 2   len  .all.      1 5.667266e-01 5.667266e-01     0.57       ns T-test three   len  .all.      2 1.371704e-05 2.743408e-05  1.4e-05     **** T-test

# Visualize ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "dose", palette = "jco")+   stat_compare_means(method = "anova", label.y = 40)+      # Add global p-value   stat_compare_means(label = "p.signif", method = "t.test",                      ref.group = ".all.")                  # Pairwise comparing against all

Add p-values too significance levels to ggplots

Influenza A virus subtype H5N1 typical situation, where pairwise comparisons against “all” tin displace move useful, is illustrated hither using the myeloma data laid upward from the survminer package.
We’ll plot the seem profile of the DEPDC1 factor according to the patients’ molecular groups. We desire to know if in that location is whatsoever departure betwixt groups. If yes, where the departure is?
To response to this question, yous tin displace perform a pairwise comparing betwixt all the 7 groups. This volition Pb to a lot of comparisons betwixt all possible combinations. If yous convey many groups, every bit here, it powerfulness move hard to interpret.
Another slow solution is to compare each of the 7 groups against “all” (i.e. base-mean). When the attempt out is significant, too then yous tin displace conclude that DEPDC1 is significantly overexpressed or downexpressed inward a grouping xxx compared to all.

# Load myeloma information from survminer bundle if(!require(survminer)) install.packages("survminer") data("myeloma", bundle = "survminer")  # Perform the attempt out compare_means(DEPDC1   molecular_group,  information = myeloma,               ref.group = ".all.", method = "t.test")

# Influenza A virus subtype H5N1 tibble: 7 x 8      .y. group1           group2            p        p.adj p.format p.signif method                                             1 DEPDC1  .all.       Cyclin D-1 1.496896e-01 4.490687e-01  0.14969       ns T-test 2 DEPDC1  .all.       Cyclin D-2 5.231428e-01 1.000000e+00  0.52314       ns T-test three DEPDC1  .all.     Hyperdiploid 2.815333e-04 1.689200e-03  0.00028      *** T-test iv DEPDC1  .all. Low os illness 5.083985e-03 2.541992e-02  0.00508       ** T-test five DEPDC1  .all.              MAF 8.610664e-02 3.444265e-01  0.08611       ns T-test half dozen DEPDC1  .all.            MMSET 5.762908e-01 1.000000e+00  0.57629       ns T-test 7 DEPDC1  .all.    Proliferation 1.241416e-09 8.689910e-09  1.2e-09     **** T-test

# Visualize the seem profile ggboxplot(myeloma, x = "molecular_group", y = "DEPDC1", color = "molecular_group",            add together = "jitter", legend = "none") +   rotate_x_text(angle = 45)+   geom_hline(yintercept = mean(myeloma$DEPDC1), linetype = 2)+ # Add horizontal occupation at base of operations hateful   stat_compare_means(method = "anova", label.y = 1600)+        # Add global annova p-value   stat_compare_means(label = "p.signif", method = "t.test",                      ref.group = ".all.")                      # Pairwise comparing against all

Add p-values too significance levels to ggplots

From the plot above, nosotros tin displace conclude that DEPDC1 is significantly overexpressed inward proliferation grouping and, it’s significantly downexpressed inward Hyperdiploid too Low os illness compared to all.

Note that, if yous desire to cover the ns symbol, specify the argument hide.ns = TRUE.

# Visualize the seem profile ggboxplot(myeloma, x = "molecular_group", y = "DEPDC1", color = "molecular_group",            add together = "jitter", legend = "none") +   rotate_x_text(angle = 45)+   geom_hline(yintercept = mean(myeloma$DEPDC1), linetype = 2)+ # Add horizontal occupation at base of operations hateful   stat_compare_means(method = "anova", label.y = 1600)+        # Add global annova p-value   stat_compare_means(label = "p.signif", method = "t.test",                      ref.group = ".all.", hide.ns = TRUE)      # Pairwise comparing against all

Add p-values too significance levels to ggplots

Multiple grouping variables

• Two independent sample comparisons later grouping the information past times unopen to other variable:

Perform the test:

compare_means(len   supp, information = ToothGrowth,                group.by = "dose")

# Influenza A virus subtype H5N1 tibble: three x ix    dose   .y. group1 group2           p      p.adj p.format p.signif   method                                  1   0.5   len     OJ     VC 0.023186427 0.04637285    0.023        * Wilcoxon 2   1.0   len     OJ     VC 0.004030367 0.01209110    0.004       ** Wilcoxon three   2.0   len     OJ     VC 1.000000000 1.00000000    1.000       ns Wilcoxon

In the illustration above, for each marker of the variable “dose”, nosotros compare the agency of the variable “len” inward the dissimilar groups formed past times the grouping variable “supp”.

Visualize (1/2). Create a multi-panel box plots facetted past times grouping (here, “dose”):

# Box plot facetted past times "dose" p <- ggboxplot(ToothGrowth, x = "supp", y = "len",           color = "supp", palette = "jco",           add together = "jitter",           facet.by = "dose", short.panel.labs = FALSE) # Use alone p.format every bit label. Remove method name. p + stat_compare_means(label = "p.format")

Add p-values too significance levels to ggplots

# Or occupation significance symbol every bit label p + stat_compare_means(label =  "p.signif", label.x = 1.5)

Add p-values too significance levels to ggplots

To cover the ‘ns’ symbol, occupation the argument hide.ns = TRUE.

Visualize (2/2). Create ane unmarried panel amongst all box plots. Plot y = “len” past times x = “dose” too color past times “supp”:

p <- ggboxplot(ToothGrowth, x = "dose", y = "len",           color = "supp", palette = "jco",           add together = "jitter") p + stat_compare_means(aes(group = supp))

Add p-values too significance levels to ggplots

# Show alone p-value p + stat_compare_means(aes(group = supp), label = "p.format")

Add p-values too significance levels to ggplots

# Use significance symbol every bit label p + stat_compare_means(aes(group = supp), label = "p.signif")

Add p-values too significance levels to ggplots

• Paired sample comparisons later grouping the information past times unopen to other variable:

Perform the test:

compare_means(len   supp, information = ToothGrowth,                group.by = "dose", paired = TRUE)

# Influenza A virus subtype H5N1 tibble: three x ix    dose   .y. group1 group2          p      p.adj p.format p.signif   method                                 1   0.5   len     OJ     VC 0.03296938 0.06593876    0.033        * Wilcoxon 2   1.0   len     OJ     VC 0.01905889 0.05717667    0.019        * Wilcoxon three   2.0   len     OJ     VC 1.00000000 1.00000000    1.000       ns Wilcoxon

Visualize. Create a multi-panel box plots facetted past times grouping (here, “dose”):

# Box plot facetted past times "dose" p <- ggpaired(ToothGrowth, x = "supp", y = "len",           color = "supp", palette = "jco",            line.color = "gray", line.size = 0.4,           facet.by = "dose", short.panel.labs = FALSE) # Use alone p.format every bit label. Remove method name. p + stat_compare_means(label = "p.format", paired = TRUE)

Add p-values too significance levels to ggplots

Other plot types

• Bar too occupation plots (one grouping variable):

# Bar plot of hateful +/-se ggbarplot(ToothGrowth, x = "dose", y = "len", add together = "mean_se")+   stat_compare_means() +                                         # Global p-value   stat_compare_means(ref.group = "0.5", label = "p.signif",                      label.y = c(22, 29))                   # compare to ref.group  # Line plot of hateful +/-se ggline(ToothGrowth, x = "dose", y = "len", add together = "mean_se")+   stat_compare_means() +                                         # Global p-value   stat_compare_means(ref.group = "0.5", label = "p.signif",                      label.y = c(22, 29))     

Add p-values too significance levels to ggplots

• Bar too occupation plots (two grouping variables):

ggbarplot(ToothGrowth, x = "dose", y = "len", add together = "mean_se",           color = "supp", palette = "jco",            seat = position_dodge(0.8))+   stat_compare_means(aes(group = supp), label = "p.signif", label.y = 29)  ggline(ToothGrowth, x = "dose", y = "len", add together = "mean_se",           color = "supp", palette = "jco")+   stat_compare_means(aes(group = supp), label = "p.signif",                       label.y = c(16, 25, 29))

Add p-values too significance levels to ggplots

Infos

This analysis has been performed using R software (ver. 3.3.2) and ggpubr (ver. 0.1.3).

Source:

https://www.r-bloggers.com/add-p-values-and-significance-levels-to-ggplots/

Sumber http://engdashboard.blogspot.com/

Tweet

Jangan sampai ketinggalan postingan-postingan terbaik dari Branches of mechanical engineering: Add Together P-Values As Well As Significance Levels To Ggplots. Berlangganan melalui email sekarang juga:

Bali Attractions