Extract P-Value from Aov

Extract p-value from aov

Here:

summary(test)[[1]][["Pr(>F)"]][1]

Extract p-value from anova() from comparison of two linear models in R

You just need quotation marks:

pvalue <- anova(m2,m1)$"Pr(>F)"

You can access the second element of pvalue using normal bracket subsetting:

pvalue[2]

So in your example I believe you'll use

Coop <- anova(M6,Models5[[B]])$"Pr(>F)"[2]

(Although without access to M6, I can't be totally sure).

The str command is very useful in this sort of situation, to figure out what kind of object you're dealing with:

str(myanova$"Pr(>F)")

How to get p-value from a summary anova test in R?

If you take a look at the structrure of your result (see ?str function) you'll realize about how to use indexing operator ([) in order to get Pr(>F).

> str(result)
List of 1
 $ :Classes ‘anova’ and 'data.frame':   2 obs. of  5 variables:
  ..$ Df     : num [1:2] 2 12
  ..$ Sum Sq : num [1:2] 35.7 261.3
  ..$ Mean Sq: num [1:2] 17.9 21.8
  ..$ F value: num [1:2] 0.82 NA
  ..$ Pr(>F) : num [1:2] 0.464 NA
 - attr(*, "class")= chr [1:2] "summary.aov" "listof"

You can do it this way

> result[[1]][["Pr(>F)"]][1]
[1] 0.4636054

or

> result[[1]][[5]][1]
[1] 0.4636054

How to extract p values from a vector with multiple anova

You can use sapply to iterate over each anova and extract p-value.

sapply(w3, function(x) summary(x)[[1]][["Pr(>F)"]][[1]])

---

Using reproducible example with mtcars

w3 <- lapply(split(mtcars, mtcars$cyl), aov, formula=mpg ~ am) 
sapply(w3, function(x) summary(x)[[1]][["Pr(>F)"]][[1]])

#     4      6      8 
#0.0892 0.2209 0.8662 

R- One way anova extracting p_value

Consider using broom. With tidy(), you can extract only the p.value field:

require(broom)
a <- aov(mpg ~ wt, mtcars)

tidy(a)
#        term df    sumsq     meansq statistic      p.value
# 1        wt  1 847.7252 847.725250  91.37533 1.293959e-10
# 2 Residuals 30 278.3219   9.277398        NA           NA

tidy(a)$p.value
# [1] 1.293959e-10           NA

How to extract p-values at a time and save p-values as csv files?

You could apply anova on each group and extract p-value from them

vals <- sapply(split(df, df$GeneSymbol), function(i) 
              anova(lm(Value ~ Label, data = i))$"Pr(>F)"[1])
vals

#       A         B         C 
#0.6419426 0.9446151 0.9146334 

If you want to write it in csv, you could do

p_data <- data.frame(p_value = vals)
write.csv(p_data, "/path/of/the/file.csv", row.names = FALSE)

---

Similarly with dplyr you could do

df %>%
  group_split(GeneSymbol) %>%
  purrr::map_dbl(~anova(lm(Value ~ Label, data = .))$"Pr(>F)"[1])

#[1] 0.6419426 0.9446151 0.9146334

data

df <- structure(list(GeneSymbol = structure(c(1L, 2L, 3L, 1L, 2L, 3L, 
1L, 2L, 3L, 1L, 2L, 3L), .Label = c("A", "B", "C"), class = "factor"), 
Value = c(0.14, 0.16, 0.01, 0.18, 0.54, 0.18, 0.2, 0.54, 
0.2, 0.02, 0.2, 0.02), Label = c(1L, 1L, 1L, 1L, 1L, 1L, 
0L, 0L, 0L, 0L, 0L, 0L)), class = "data.frame", row.names = c("2", 
"3", "4", "5", "6", "7", "8", "9", "10", "11", "12", "13"))

How to use lapply() with aov() because results returns no p-value and f-value but just a list of Class and Mode ?

I'm not sure your purpose applying those tests.
If code below is your purpose,

    cary <- aov(df_list[["1.3.A"]]$Height ~ df_list[["1.3.A"]]$Person)
    summary(cary)

and kruskal wallis test that gives you p-value was

    lapply(df_list, kruskal.test)

then I'm not sure p-value from code above is correct.
That code will print warning message that
In kruskal.test.default(X[[i]], ...) :
some elements of 'x' are not numeric and will be coerced to numeric

This coerced person's name as numeric.
It's pretty messy but code below may be messy but gives better p-values

    lapply(df_list, function(x) (kruskal.test(x$Height ~ x$Person)))
    lapply(df_list, function(x) summary(aov(x$Height ~ x$Person)))

EDIT:
If you want to get only p-values,

    lapply(df_list, function(x) (kruskal.test(x$Height ~ x$Person)$p.value))
    lapply(df_list, function(x) summary(aov(x$Height ~ x$Person))[[1]]$`Pr(>F)`[1])

will do.

EDIT 2(reply to comments)

1. If you want your tree function work without using $, then you should change code as

   tree <- function(df) {   aov(Height ~ Person, data = df) } 
   

or

    tree <- function(df) {   aov(df$Height ~ df$Person) } 

and then, to print p-value, instead of using

    summary(shrub)

use

    lapply(shrub,summary)

2. Because I use lapply, it fits ANOVA for each data frame of list.
   To call p-values from summary result, try fitting code below

    x<-summary(aov(Height~Person, df_list[[1]]))
   

then you'll see x is List of 1 and to call components of x and because there is only one element in x, so i call with [[1]]

And then, x has form of matrix so with

    x[[1]]$`Pr(>F)`

it will print

    [1] 1.433592e-42           NA

I want to remove NA so i add [1] to the end like

    x[[1]]$`Pr(>F)`[1]

---

Related Topics