Adding x and y variable labels in corrplot

KJcatmaster · November 5, 2024, 9:38am

Hi

I’m new to this! Please help

I have correlation values calculated from a previous genomic analysis that I’m trying to insert into corrplot to produce a heatmap. I have managed to get my heat map however for some reason I am struggling to show the labels for each block on the y and x axis.

# Load necessary libraries
library(readxl)  # Load the readxl package
library(corrplot)  # Load the corrplot package

# Read the Excel file
data <- read_excel("heatmap_correlations.xlsx", col_names = TRUE)

# Convert tibble to data frame
data <- as.data.frame(data)

# Check for duplicate entries in the first column and make them unique if necessary
if (any(duplicated(data[, 1]))) {
  warning("Duplicate row names found. They will be made unique.")
  data[, 1] <- make.unique(as.character(data[, 1]))
}

# Set the first column as row names and remove it from the data frame
colnames(data) <- data[, 1]
data <- data[, -1]

# Remove any empty rows or columns (if applicable)
data <- na.omit(data)
data <- data[rowSums(is.na(data)) != ncol(data), colSums(is.na(data)) != nrow(data)]

# Convert all columns to numeric
data[] <- lapply(data, function(x) as.numeric(as.character(x)))

# Check for any non-numeric columns after conversion
if (any(is.na(data))) {
  stop("Non-numeric values found in the data. Please check the Excel file.")
}

# Generate the correlation matrix
cor_matrix <- cor(data, use = "pairwise.complete.obs")

# Create the correlation plot
corrplot(cor_matrix, method = "circle", 
         col = colorRampPalette(c("turquoise", "white", "orange"))(400), 
         type = "lower", 
         order = "hclust", 
         addgrid.col = "grey", 
         tl.col = "black", 
         tl.pos = 'n',
         tl.srt = 40, 
         tl.offset = 1,
         number.cex = 0.4,
         addCoef.col = "black",
         title = "Genetic Correlation", cex.main = 1)
(mar=c(5.1,4.1,4.1,2.1))

FJCC · November 5, 2024, 1:07pm

Setting t.plot = 'n' prevents the plotting of the labels. Here's a demonstration:

MAT <- matrix(rnorm(25), nrow = 5)
dimnames(MAT) <- list(c("A","B","C","D","E"), c("V","W","X","Y","Z"))
corrplot::corrplot(cor(MAT))

corrplot::corrplot(cor(MAT), tl.pos = "n")

^{Created on 2024-11-05 with reprex v2.1.1}

KJcatmaster · November 5, 2024, 8:42pm

Ah, simple fix! I managed to get the labels in, however its not giving the corresponding labels as I have it in my data frame - its not a case of just renaming the labels, I have 17 rows and columns so I am trying to figure out what its doing

FJCC · November 6, 2024, 1:24am

Please post some or all of your data. For the whole data set, post the output of

dput(data)

Use the version of data that is used in corrplot(). If the data set is large, you can post the first 25 rows with

dput(head(data, 25))

KJcatmaster · November 6, 2024, 6:02am

Thank you! here it is

structure(list(...1 = c("All", "GGE", "FE", "GTCS", "CAE", "JME", 
"JAE", "F_HS", "F_Neg", "F_Oth", "ACC", "AMY", "BRS", "CAU", 
"PALL", "PUT", "THA"), All = c("0", "0.9241", "0.8865", "0.9899", 
"0.983", "0.8471", "0.9437", "0.6045", "0.7427", "0.705", "-0.0324", 
"-0.0915", "0.0746", "-0.0454", "-0.1761", "-0.0641", "-0.0726"
), GGE = c("0.9241", "0", "0.6148", "0.9451", "1.046", "0.8864", 
"0.9451", "0.2281", "0.46", "0.7098", "-0.0449", "-0.165", "-0.0942", 
"-0.0593", "-0.1312", "-0.0077", "-0.0564"), FE = c("0.8865", 
"0.6148", "0", "0.7344", "0.7326", "0.5694", "0.67", "0.8983", 
"0.8853", "0.6495", "0.0049", "0.096", "-0.0627", "0.0429", "-0.2017", 
"-0.0.0847", "-0.0646"), GTCS = c("0.9899", "0.9451", "0.7344", 
"0", "1.1472", "0.7239", "0.7395", "-0.267", "0.5303", "0.987", 
"-0.239", "-0.3777", "-0.0728", "-0.1326", "-0.1355", "-0.0546", 
"-0.068"), CAE = c("0.983", "1.046", "0.7326", "1.1472", "0", 
"0.7907", "1.0362", "0.2271", "0.4703", "1.0508", "-0.0542", 
"-0.1365", "-0.1027", "-0.065", "-0.1829", "-0.0709", "-0.0544"
), JME = c("0.8471", "0.8864", "0.5694", "0.7239", "0.7907", 
"0", "0.6832", "0.2247", "0.4481", "0.8412", "-0.0426", "-0.2074", 
"-0.1544", "-0.0038", "-0.1703", "0.0105", "-0.1299"), JAE = c("0.9437", 
"0.9451", "0.67", "0.7395", "1.0362", "0.6832", "0", "0.2063", 
"0.4054", "1.341", "-0.0798", "-0.1141", "0.0473", "-0.1685", 
"-0.1764", "-0.0656", "-0.0335"), F_HS = c("0.6045", "0.2281", 
"0.8983", "-0.267", "0.2271", "0.2247", "0.2063", "0", "0.5171", 
"2.0792", "0.1217", "-0.118", "-0.216", "-0.0125", "-0.0708", 
"-0.0096", "-0.0714"), F_Neg = c("0.7427", "0.46", "0.8853", 
"0.5303", "0.4703", "0.4481", "0.4054", "0.5171", "0", "0.5303", 
"0.0871", "0.0395", "-0.0799", "0.0321", "-0.0885", "0.0142", 
"-0.0943"), F_Oth = c("0.705", "0.7098", "0.6495", "0.987", "1.0508", 
"0.8412", "1.341", "2.0792", "0.5303", "0", "-0.2197", "-0.201", 
"-0.0471", "0.1705", "-0.366", "-0.3629", "-0.1624"), ACC = c("-0.0324", 
"-0.0449", "0.0049", "-0.239", "-0.0542", "-0.0426", "-0.0798", 
"0.1217", "0.0871", "-0.2197", "0", "0.3194", "0.0654", "0.4447", 
"0.2815", "0.4359", "0.2733"), AMY = c("-0.0915", "-0.165", "0.096", 
"-0.3777", "-0.1365", "-0.2074", "-0.1141", "-0.118", "0.0395", 
"-0.201", "0.3194", "0", "0.0545", "0.1233", "0.1622", "0.2413", 
"0.2681"), BRS = c("0.0746", "-0.0942", "-0.0627", "-0.0728", 
"-0.1027", "-0.1544", "0.0473", "-0.216", "-0.0799", "-0.0471", 
"0.0654", "0.0545", "0", "0.098", "0.4803", "0.1129", "0.4972"
), CAU = c("-0.0454", "-0.0593", "0.0429", "-0.1326", "-0.065", 
"-0.0038", "-0.1685", "-0.0125", "0.0321", "0.1705", "0.4447", 
"0.1233", "0.098", "0", "0.3486", "0.508", "0.0764"), PALL = c("-0.1761", 
"-0.1312", "-0.2017", "-0.1355", "-0.1829", "-0.1703", "-0.1764", 
"-0.0708", "-0.0885", "-0.366", "0.2815", "0.1622", "0.4803", 
"0.3486", "0", "0.5521", "0.3878"), PUT = c("-0.0641", "-0.0077", 
"-0.0.0847", "-0.0546", "-0.0709", "0.0105", "-0.0656", "-0.0096", 
"0.0142", "-0.3629", "0.4359", "0.2413", "0.1129", "0.508", "0.5521", 
"0", "0.2192"), THA = c("-0.0726", "-0.0564", "-0.0646", "-0.068", 
"-0.0544", "-0.1299", "-0.0335", "-0.0714", "-0.0943", "-0.1624", 
"0.2733", "0.2681", "0.4972", "0.0764", "0.3878", "0.2192", "0"
)), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, 
-17L))

FJCC · November 6, 2024, 2:36pm

I'm not sure what the data is that you posted. It can't be what is passed to corrplot() because all the values are text (they are wrapped in double quotes), the first column appears to be row names, and some of the values are outside of the range [-1, 1]. I manipulated the data so I could produce a correlation plot of its columns and that plot has the proper labels. If that does not help you get the result you need, please post the data from the object that produced the plot you showed previously with the labels that started with NA..

data <- structure(list(...1 = c("All", "GGE", "FE", "GTCS", "CAE", "JME", 
                                "JAE", "F_HS", "F_Neg", "F_Oth", "ACC", "AMY", "BRS", "CAU", 
                                "PALL", "PUT", "THA"), 
                       All = c("0", "0.9241", "0.8865", "0.9899", 
                               "0.983", "0.8471", "0.9437", "0.6045", "0.7427", "0.705", "-0.0324", 
                               "-0.0915", "0.0746", "-0.0454", "-0.1761", "-0.0641", "-0.0726"
                       ), 
                       GGE = c("0.9241", "0", "0.6148", "0.9451", "1.046", "0.8864", 
                               "0.9451", "0.2281", "0.46", "0.7098", "-0.0449", "-0.165", "-0.0942", 
                               "-0.0593", "-0.1312", "-0.0077", "-0.0564"), 
                       FE = c("0.8865", "0.6148", "0", "0.7344", "0.7326", "0.5694", "0.67", "0.8983", 
                              "0.8853", "0.6495", "0.0049", "0.096", "-0.0627", "0.0429", "-0.2017", 
                              "-0.0.0847", "-0.0646"), 
                       GTCS = c("0.9899", "0.9451", "0.7344", 
                                "0", "1.1472", "0.7239", "0.7395", "-0.267", "0.5303", "0.987", 
                                "-0.239", "-0.3777", "-0.0728", "-0.1326", "-0.1355", "-0.0546", 
                                "-0.068"), 
                       CAE = c("0.983", "1.046", "0.7326", "1.1472", "0", 
                               "0.7907", "1.0362", "0.2271", "0.4703", "1.0508", "-0.0542", 
                               "-0.1365", "-0.1027", "-0.065", "-0.1829", "-0.0709", "-0.0544"
                       ), 
                       JME = c("0.8471", "0.8864", "0.5694", "0.7239", "0.7907", 
                               "0", "0.6832", "0.2247", "0.4481", "0.8412", "-0.0426", "-0.2074", 
                               "-0.1544", "-0.0038", "-0.1703", "0.0105", "-0.1299"), 
                       JAE = c("0.9437", "0.9451", "0.67", "0.7395", "1.0362", "0.6832", "0", "0.2063", 
                               "0.4054", "1.341", "-0.0798", "-0.1141", "0.0473", "-0.1685", 
                               "-0.1764", "-0.0656", "-0.0335"), 
                       F_HS = c("0.6045", "0.2281", "0.8983", "-0.267", "0.2271", "0.2247", "0.2063", "0", "0.5171", 
                                "2.0792", "0.1217", "-0.118", "-0.216", "-0.0125", "-0.0708", 
                                "-0.0096", "-0.0714"), 
                       F_Neg = c("0.7427", "0.46", "0.8853", "0.5303", "0.4703", "0.4481", "0.4054", "0.5171", "0", "0.5303", 
                                 "0.0871", "0.0395", "-0.0799", "0.0321", "-0.0885", "0.0142", 
                                 "-0.0943"), 
                       F_Oth = c("0.705", "0.7098", "0.6495", "0.987", "1.0508", 
                                 "0.8412", "1.341", "2.0792", "0.5303", "0", "-0.2197", "-0.201", 
                                 "-0.0471", "0.1705", "-0.366", "-0.3629", "-0.1624"), 
                       ACC = c("-0.0324", "-0.0449", "0.0049", "-0.239", "-0.0542", "-0.0426", "-0.0798", 
                               "0.1217", "0.0871", "-0.2197", "0", "0.3194", "0.0654", "0.4447", 
                               "0.2815", "0.4359", "0.2733"), 
                       AMY = c("-0.0915", "-0.165", "0.096", "-0.3777", "-0.1365", "-0.2074", "-0.1141", "-0.118", "0.0395", 
                               "-0.201", "0.3194", "0", "0.0545", "0.1233", "0.1622", "0.2413", 
                               "0.2681"), 
                       BRS = c("0.0746", "-0.0942", "-0.0627", "-0.0728", 
                               "-0.1027", "-0.1544", "0.0473", "-0.216", "-0.0799", "-0.0471", 
                               "0.0654", "0.0545", "0", "0.098", "0.4803", "0.1129", "0.4972"
                       ), 
                       CAU = c("-0.0454", "-0.0593", "0.0429", "-0.1326", "-0.065", 
                               "-0.0038", "-0.1685", "-0.0125", "0.0321", "0.1705", "0.4447", 
                               "0.1233", "0.098", "0", "0.3486", "0.508", "0.0764"), 
                       PALL = c("-0.1761", "-0.1312", "-0.2017", "-0.1355", "-0.1829", "-0.1703", "-0.1764", 
                                "-0.0708", "-0.0885", "-0.366", "0.2815", "0.1622", "0.4803", 
                                "0.3486", "0", "0.5521", "0.3878"), 
                       PUT = c("-0.0641", "-0.0077", "-0.0.0847", "-0.0546", "-0.0709", "0.0105", "-0.0656", "-0.0096", 
                               "0.0142", "-0.3629", "0.4359", "0.2413", "0.1129", "0.508", "0.5521", 
                               "0", "0.2192"), 
                       THA = c("-0.0726", "-0.0564", "-0.0646", "-0.068", "-0.0544", "-0.1299", "-0.0335", "-0.0714", "-0.0943", "-0.1624", 
                               "0.2733", "0.2681", "0.4972", "0.0764", "0.3878", "0.2192", "0"
                               
                       )), class = c("tbl_df", "tbl", "data.frame"), row.names = c(NA, -17L))
library(corrplot)
#> corrplot 0.94 loaded
library(tidyverse)
data2 <- data |> select(-1) |> mutate(across(.col = 1:17, as.numeric))
#> Warning: There were 2 warnings in `mutate()`.
#> The first warning was:
#> ℹ In argument: `across(.col = 1:17, as.numeric)`.
#> Caused by warning:
#> ! NAs introduced by coercion
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 1 remaining warning.
cor_matrix2 <- cor(data2, use = "complete.obs")

corrplot(cor_matrix2, method = "circle", 
         col = colorRampPalette(c("turquoise", "white", "orange"))(400), 
         type = "lower", 
         order = "hclust", 
         addgrid.col = "grey", 
         tl.col = "black", 
         tl.srt = 40, 
         tl.offset = 1,
         number.cex = 0.4,
         addCoef.col = "black",
         title = "Genetic Correlation", cex.main = 1)
#> Warning in ind1:ind2: numerical expression has 2 elements: only the first used

^{Created on 2024-11-06 with reprex v2.1.1}

KJcatmaster · November 6, 2024, 6:48pm

It worked! Thank you so much, you are a lifesaver! I noticed that the diagonal labels are shifted up, even when I had originally added the labels (which added as 'NA'), so they do no correspond to the relative block, i.e is there a way I am able to shift only the diagonal labels one block down? (So that the first label "CAU" is to the right of the top block instead of an above it? Then the F_Neg would end up being to the right of the bottom block?

FJCC · November 6, 2024, 10:42pm

I don't know how to change the position of the labels on the diagonal and, in any case, it looks correct to me. Those labels are for the columns of the figure and they seem reasonably aligned.