R help - Jul 2024 - Reinterpret data without saving it to a file 1st? Check for integer stopping at 1st decimal?

A small number of columns in the data I need to work with are strings, the
rest numbers.  I'm using read_excel() from the readxl package to get the
data ; right after it, the string columns are of type chr and the rest num.
I'm tasked with finding out which columns are integers. From an advice, I
tried saving the spreadsheet content into a CSV then loading that, which
works like a charm ; the chr columns are the same but now a large portion
of num is now instead int. Is there a way to skip writing and reading a CSV
and get the same transformation? Perhaps some way to break the spreadsheet
data (eg XLdata <- read_excel(...)), then put it back together without any
writing to a file (eg XLdataReformed <- reform(XLdata)) ?

In addition, from is.integer() documentation I ran
> is.wholenumber <- function(x, tol = .Machine$double.eps^0.5) abs(x -
round
> (x)) < tol
and I'm now trying to have it stop at the 1st decimal content of a column.
Someone advised me to use break and I scripted
> is_integer = TRUE for (current_row in seq_along(data$column)) { if (!
> is.wholenumber(data$column[current_row])) { is_integer = FALSE break; } }
but I'm wondering if there's something better to check if a column is
entirely made of integers.

Thank you kindly for your help

	[[alternative HTML version deleted]]

? Sun, 14 Jul 2024 03:16:56 -0400
DynV Montrealer <dynvec at gmail.com> ?????:
> Perhaps some way to break the spreadsheet data (eg XLdata <-
> read_excel(...)), then put it back together without any writing to a
> file (eg XLdataReformed <- reform(XLdata)) ?
read_excel() is documented to return objects of class tibble:
https://cran.r-project.org/package=tibble/vignettes/tibble.html

Long story short, tibbles are named lists of columns, so it should be
possible for you to access and replace the individual parts of them
using the standard list subset syntax XLdata[[columnname]].

Lists are described in R Intro chapter 6 and many other books on R:
https://cran.r-project.org/doc/manuals/R-intro.html#Lists-and-data-frames
http://web.archive.org/web/20230415001551if_/http://ashipunov.info/shipunov/school/biol_240/en/visual_statistics.pdf
(see section 3.8.2 on page 93 and following)
> In addition, from is.integer() documentation I ran
> 
> > is.wholenumber <- function(x, tol = .Machine$double.eps^0.5) abs(x
> > - round (x)) < tol  
> 
> and I'm now trying to have it stop at the 1st decimal content of a
> column.
If you'd like to write idiomatic R code, consider the fact that
is.wholenumber is vectorised:

is.wholenumber(c(1,2,3,pi))
# [1]  TRUE  TRUE  TRUE FALSE

Given a vector of numbers, it will return a vector of the same length
specifying whether each element can be considered a whole number.
Combine it with all() and you can test the whole column in two function
calls.

R also has a type.convert function that may be useful in this case:
https://search.r-project.org/R/refmans/utils/html/type.convert.html

-- 
Best regards,
Ivan

The answer:
https://statisticsglobe.com/change-classes-data-frame-columns-automatically-r

On Sun, Jul 14, 2024 at 3:16?AM DynV Montrealer <dynvec at gmail.com>
wrote:
> A small number of columns in the data I need to work with are strings, the
> rest numbers.  I'm using read_excel() from the readxl package to get
the
> data ; right after it, the string columns are of type chr and the rest num.
> I'm tasked with finding out which columns are integers. From an advice,
I
> tried saving the spreadsheet content into a CSV then loading that, which
> works like a charm ; the chr columns are the same but now a large portion
> of num is now instead int. Is there a way to skip writing and reading a CSV
> and get the same transformation? Perhaps some way to break the spreadsheet
> data (eg XLdata <- read_excel(...)), then put it back together without
any
> writing to a file (eg XLdataReformed <- reform(XLdata)) ?
>
> In addition, from is.integer() documentation I ran
>
> > is.wholenumber <- function(x, tol = .Machine$double.eps^0.5) abs(x
-
> round
> > (x)) < tol
>
> and I'm now trying to have it stop at the 1st decimal content of a
column.
> Someone advised me to use break and I scripted
>
> > is_integer = TRUE for (current_row in seq_along(data$column)) { if (!
> > is.wholenumber(data$column[current_row])) { is_integer = FALSE break;
} }
>
> but I'm wondering if there's something better to check if a column
is
> entirely made of integers.
>
> Thank you kindly for your help
>
>
	[[alternative HTML version deleted]]