Home / Questions / 345753
Accepted
Juan Luis
Asked: 2020-04-14 12:48:07 +0800 CST 2020-04-14 12:48:07 +0800 CST

How to display the descriptive values ​​in a certain way for a binomial logit model and what other package to use to display the data?

  • 772

I'm trying to build a styled table in descriptive analysis Rmarkdown for my binomial logit model .

descr For now, I use the library command summarytoolsand library printras follows (with my own data.frame):

```{r , eval=TRUE, echo=FALSE}
library(printr)
df<-mtcars
library(summarytools)
descr(datos_modelo, stats = c("mean", "sd", "max", "min"), na.rm =TRUE)

The problem is that the table that is generated for me puts the descriptive values ​​in the columns and the variables by rows. But I have many variables. so it would be more correct if the descriptions were in the rows and the variables in the columns.

How can I "transpose" this table?

Knowing that I am developing a binomial logit with the package glmwith categorical variables (factor type so that R understands me) both for the endogenous variable and for most of the exogenous variables , is there any recommended package specifically for this type of descriptive analysis? I see that I have many packages available to me but I don't know which one is the most suitable for my work.

r

1 Answers

  1. Best Answer

    Imagining you have something like this:

    descr(mtcars, stats = c("mean", "sd", "max", "min"))

Descriptive Statistics  
mtcars  
N: 32  

                  am   carb    cyl     disp   drat   gear       hp     mpg    qsec     vs     wt
------------- ------ ------ ------ -------- ------ ------ -------- ------- ------- ------ ------
         Mean   0.41   2.81   6.19   230.72   3.60   3.69   146.69   20.09   17.85   0.44   3.22
      Std.Dev   0.50   1.62   1.79   123.94   0.53   0.74    68.56    6.03    1.79   0.50   0.98
          Max   1.00   8.00   8.00   472.00   4.93   5.00   335.00   33.90   22.90   1.00   5.42
          Min   0.00   1.00   4.00    71.10   2.76   3.00    52.00   10.40   14.50   0.00   1.51

    You can use the traditional "transpose" of arrays, t()this because the output of descr()is an object that inherits from the class matrix:

    t(descr(mtcars, stats = c("mean", "sd", "max", "min")))
Descriptive Statistics  
mtcars  
N: 32  

               Mean   Std.Dev      Max     Min
---------- -------- --------- -------- -------
        am     0.41      0.50     1.00    0.00
      carb     2.81      1.62     8.00    1.00
       cyl     6.19      1.79     8.00    4.00
      disp   230.72    123.94   472.00   71.10
      drat     3.60      0.53     4.93    2.76
      gear     3.69      0.74     5.00    3.00
        hp   146.69     68.56   335.00   52.00
       mpg    20.09      6.03    33.90   10.40
      qsec    17.85      1.79    22.90   14.50
        vs     0.44      0.50     1.00    0.00
        wt     3.22      0.98     5.42    1.51

class(descr(mtcars, stats = c("mean", "sd", "max", "min")))
[1] "summarytools" "matrix"

    And finally to keep the original order of the columns, you can do this:

    new.df <- t(descr(mtcars, stats = c("mean", "sd", "max", "min"), na.rm =TRUE))
new.df[order(match(rownames(new.df), colnames(mtcars))),]

          Mean     Std.Dev     Max    Min
mpg   20.090625   6.0269481  33.900 10.400
cyl    6.187500   1.7859216   8.000  4.000
disp 230.721875 123.9386938 472.000 71.100
hp   146.687500  68.5628685 335.000 52.000
drat   3.596563   0.5346787   4.930  2.760
wt     3.217250   0.9784574   5.424  1.513
qsec  17.848750   1.7869432  22.900 14.500
vs     0.437500   0.5040161   1.000  0.000
am     0.406250   0.4989909   1.000  0.000
gear   3.687500   0.7378041   5.000  3.000
carb   2.812500   1.6152000   8.000  1.000

    With order(match(rownames(new.df), colnames(mtcars)))we redo the order of the data.frameoriginal, the only drawback is that you end up with a flat array and not an object of type summarytools.

    • 1