CODEWITHSUNDEEP
rdataframe
Vishesh Shrivastav
Vishesh Shrivastav

Reputation: 2129

Add calculated column to dataframe R

I want to calculate some statistics (mean, min, max, standard deviation etc) of some columns from a dataframe and store these values as another dataframe.

Here's a sample:

>foo

    Col1 Col2 Col3 Col4
1    1    6   10   60
2    2    7   20   70
3    3    8   30   80
4    4    9   40   90
5    5   10   50  100

For example, I want to store the mean and minimum value of Col1 and Col3 in a dataframe like this:

>bar

           Col1       Col3
Mean          3         30
Min           1         10

I want to do this through a loop, something like this:

# Result dataframe
bar <- data.frame(Col1 = integer(), Col3 = integer())

variables_for_stats <- c("Col1","Col3")

# I want to do something on the lines of this:
for (z in variables_for_stats){

    # Populate column with required values
    col <- c(mean(foo$z,min(foo$z)) # Throws an error - argument is not numeric or logical: returning NA

    # Add col to 'bar'
    bar$z<- col # Does not work
}

My actual foo dataframe currently has ~ 40 columns and actual variables_for_stats are around 20. Both of these can change, hence the desire to do this through a for loop and lists. How should I do this ?

Upvotes: 4

Views: 6248

Answers (4)

giocomai
giocomai

Reputation: 3518

if you're interested in a tidyverse solution...

library(tidyverse)

foo <- tribble(~Col1, ~Col2, ~Col3, ~Col4,
               1,    6,   10,   60,
               2,    7,   20,   70,
               3,    8,   30,   80,
               4,    9,   40,   90,
               5,   10,   50,  100)

foo %>%
  gather(Col, Value) %>% 
  group_by(Col) %>% 
  summarise(Mean = mean(Value), Minimum = min(Value))
#> # A tibble: 4 x 3
#>   Col    Mean Minimum
#>   <chr> <dbl>   <dbl>
#> 1 Col1      3       1
#> 2 Col2      8       6
#> 3 Col3     30      10
#> 4 Col4     80      60

Edit If you want the resulting data frame exactly as you pointed out in your question, then:

foo %>%
  gather(Col, Value) %>% 
  group_by(Col) %>% 
  summarise(Mean = mean(Value),
            Minimum = min(Value)) %>% 
  gather(Func, Value, 2:3) %>% 
  spread(Col, Value) %>% 
  select(Func, Col1, Col3)

# A tibble: 2 x 3
#  Func     Col1  Col3
#  <chr>   <dbl> <dbl>
#1 Mean        3    30
#2 Minimum     1    10

Upvotes: 3

Onyambu
Onyambu

Reputation: 79208

Using base R, you can do something like:

aggregate( values~ind,stack(foo),function(x)
     c(mean=mean(x),sd=sd(x),min=min(x),max=max(x)))#Write all the functions you want
   ind values.mean  values.sd values.min values.max
1 Col1    3.000000   1.581139   1.000000   5.000000
2 Col2    8.000000   1.581139   6.000000  10.000000
3 Col3   30.000000  15.811388  10.000000  50.000000
4 Col4   80.000000  15.811388  60.000000 100.000000

If at all you only need the summary statistics then: 

 library(tidyverse)
 summary(foo)%>%
     data.frame()%>%
     select(-Var1)%>%
     separate(Freq,c("Fun","Val"),":")%>%
     spread(Fun, Val)

       Var2 1st Qu. 3rd Qu. Max.    Mean    Median  Min.   
1      Col1     2       4       5       3       3       1  
2      Col2     7       9      10       8       8       6  
3      Col3    20      40      50      30      30      10  
4      Col4    70      90     100      80      80      60

Upvotes: 2

Calum You
Calum You

Reputation: 15062

You can do this with tidyverse tools. The actual calculation is just the summarise, the rest is just to convert the output into your desired format.

library(tidyverse)
foo <- read_table2(
  "Col1 Col2 Col3 Col4
1    6   10   60
2    7   20   70
3    8   30   80
4    9   40   90
5   10   50  10"
)

bar <- foo %>%
  summarise_at(
    .vars = vars(Col1, Col3),
    .funs = funs(mean, min)
  ) %>%
  gather(stat, value) %>%
  separate(stat, into = c("Col", "Func")) %>%
  spread(Col, value)
bar
#> # A tibble: 2 x 3
#>   Func   Col1  Col3
#>   <chr> <dbl> <dbl>
#> 1 mean      3    30
#> 2 min       1    10

Created on 2018-06-04 by the reprex package (v0.2.0).

Upvotes: 1

akrun
akrun

Reputation: 887028

We can loop over the columns of interest and get the mean and min

sapply(foo[c('Col1', 'Col3')], function(x) c(Mean = mean(x), Min =min(x)))
#      Col1 Col3
#Mean    3   30
#Min     1   10

NOTE: An apply based solution is a loop as well. But, it gives more control than a for loop in understanding the output

Upvotes: 3

Related Questions