Why do I get the message "'names' attribute must be the same length as vector" when I run a for loop with ANOVA (R)?

Question

I'm trying to create a for loop that runs ANOVA. I have a list of dataframes for which I need to run ANOVA, and I want to create a list with the resulting p-values. However, when I run the for loop, I get the following message:

'names' attribute [5] must be the same length as the vector [3]

This is the code I used:

##split my dataframe into a list of dataframes based on what gene the data represents
gene_data_list <- split(df_test, f = df_test$Gene)

##create an empty list to collect p-values
p_values <- list()

##run ANOVA on the list of dataframes
for(i in 1:length(gene_data_list)) {
  anova <- aov(value ~ variable, data = gene_data_list[[i]])
  summary <- anova_summary(anova)
  append(p_values, summary$p)
}

When I run the same code on gene_data_list[[1]] outside the loop, it runs fine. I've attached a link to the dataframe I've been using: https://easyupload.io/z5avik

Answer 1

You can do all this with slightly simpler coding using lapply instead of a for loop:

df_test <- read.csv("Test_dataset.csv")
gene_data_list <- split(df_test, f = df_test$Gene)

# Run ANOVAs
gene_anova <- lapply(gene_data_list, function(x) aov(value ~ variable, data = x))

# Get summaries
summary_list <- lapply(gene_anova, rstatix::anova_summary)

# Create single table
results_table <- do.call(rbind, summary_list)

Output:

head(results_table)

# Effect DFn DFd          F     p p<.05     ges
# 1          1   2  24 1.3570e+00 0.276     1 0.10200
# AAGAB      1   2  15 1.9334e+27 0.000     2 1.00000
# AAK1       1   2  42 4.0000e-03 0.997     1 0.00017
# AAMDC      1   2   6 4.4182e+29 0.000     2 1.00000
# AATF       1   2  24 8.5433e+28 0.000     2 1.00000
# ABCA2      1   2  33 1.0470e+00 0.362     1 0.06000

Note that if you just wanted a list of p values, you could do

p_list <- lapply(gene_anova, function(x) rstatix::anova_summary(x)$p)

Which gives you a list with each position as the p value