How to select columns with dplyr/tidyvese depending on minimal value in column in R

Question

i have a data set of certain counts of Landcoverpixel per Point.

    species_distr <- data.frame(structure(list(Point = c(101, 102, 103, 104, 105, 106), `Herbaceous cover` = c(NA_real_, 

NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), `Tree or shrub cover` = c(NA_real_, 

NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), `Cropland, irrigated or post-flooding` = c(NA_real_, 

NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), `Mosaic cropland (>50%) / natural vegetation (tree, shrub, herbaceous cover) (<50%)` = c(NA_integer_, 

NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_

), `Mosaic natural vegetation (tree, shrub, herbaceous cover) (>50%) / cropland (<50%)` = c(NA_real_, 

NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), `Tree cover, broadleaved, evergreen, closed to open (>15%)` = c(NA_integer_, 

NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_

), `Tree cover, broadleaved, deciduous, closed to open (>15%)` = c(NA_real_, 

NA_real_, NA_real_, NA_real_, NA_real_, NA_real_), `Tree cover, broadleaved, deciduous, closed (>40%)` = c(NA_integer_, 

NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_

), `Tree cover, broadleaved, deciduous, open (15-40%)` = c(NA_integer_, 

NA_integer_, NA_integer_, NA_integer_, NA_integer_, NA_integer_

), `Tree cover, needleleaved, evergreen, closed to open (>15%)` = c(NA, 

NA, 1.73725490196078, NA, NA, NA), `Tree cover, needleleaved, evergreen, closed (>40%)` = c(NA, 

NA, 0L, NA, NA, NA), `Tree cover, needleleaved, evergreen, open (15-40%)` = c(NA, 

NA, 0L, NA, NA, NA), `Tree cover, needleleaved, deciduous, closed to open (>15%)` = c(2059.57647058824, 

544, 2209.63529411765, 1226.7568627451, 1722.34901960784, 1359.10196078432

), `Tree cover, needleleaved, deciduous, closed (>40%)` = c(NA, 

NA, 0L, 0L, NA, NA), `Tree cover, needleleaved, deciduous, open (15-40%)` = c(NA, 

NA, 0L, 0L, NA, NA), `Tree cover, mixed leaf type (broadleaved and needleleaved)` = c(NA, 

NA, 1.96470588235294, 0, NA, NA), `Mosaic tree and shrub (>50%) / herbaceous cover (<50%)` = c(NA, 

NA, 0, 2, NA, NA), `Mosaic herbaceous cover (>50%) / tree and shrub (<50%)` = c(NA, 

NA, 0L, NA, NA, NA), Shrubland = c(NA, NA, 0, NA, NA, NA), `Shrubland evergreen` = c(NA, 

NA, 0L, NA, NA, NA), `Shrubland deciduous` = c(NA, NA, 0, NA, 

NA, NA), Grassland = c(NA, NA, 0L, NA, NA, NA), `Lichens and mosses` = c(NA, 

NA, 0L, NA, NA, NA), `Sparse vegetation (tree, shrub, herbaceous cover) (<15%)` = c(NA, 

NA, 0, NA, NA, NA), `Sparse tree (<15%)` = c(NA, NA, 0L, NA, 

NA, NA), `Sparse shrub (<15%)` = c(NA, NA, 0L, NA, NA, NA), `Sparse herbaceous cover (<15%)` = c(NA, 

NA, 0L, NA, NA, NA), `Tree cover, flooded, fresh or brakish water` = c(NA, 

NA, 0, NA, NA, NA), `Tree cover, flooded, saline water` = c(NA, 

NA, 0L, NA, NA, NA), `Shrub or herbaceous cover, flooded, fresh/saline/brakish water` = c(NA, 

NA, 0, NA, NA, NA), `Urban areas` = c(NA, NA, 0L, NA, NA, NA), 

    `Bare areas` = c(NA, NA, 0, NA, NA, NA), `Consolidated bare areas` = c(NA, 

    NA, 0L, NA, NA, NA), `Unconsolidated bare areas` = c(NA, 

    NA, 0L, NA, NA, NA), `Water bodies` = c(NA, NA, 4.73725490196078, 

    NA, NA, NA)), row.names = c(NA, -6L), class = c("tbl_df", 

"tbl", "data.frame")))

If want to exclude all columns that dont have a value over for example, 50. My quick and dirty solution ist this:

c <- NULL

for (i in 2:length(species_distr)) {

  if (max(na.omit(species_distr[,i])) > 50) {

    c <- c(c, i)

  }

}

species_distr_plot <- species_distr[,c(1,c)]

How can i achieve this with dplyr/tidyverse? i tried so far:

  %>%

select_if(na.omit(max(.)) > 50)

Answer 1

Here, we should use any:

library(dplyr)

species_distr %>% 

     select_if(~ !any(na.omit(max(.x)) > 50))

If you want to learn R Programming visit this R Programming Course.