Correctly convert value into scientific notation for text in R

Question

I am a bit aimless with a problem with my data frame. Below a table of an SQL server and one of the columns brings the numbers of each process. This value is a long numerical sequence that R converts to scientific notation. The problem is that when I change the column to text it brings the wrong number. How do I make R understand the correct number?

Here is a code to illustrate the problem: When I switch to text the result comes with the end 6 and not 7.

df <- data.frame(Processo = (25351001641201357))
df

> df
    Processo
   1 2.5351e+16

df1 <- as.character(df$Processo)
df1

> df1
[1] "25351001641201356"

Adding the result of the solution suggested, but which has not yet worked.

> options(scipen = 1e9)
> df <- data.frame(Processo = c(25351001641201357))
> df
           Processo
1 25351001641201356

> as.character(format(df, scientific = FALSE))
[1] "25351001641201356"

I’m using Windows 10 Home Single Language 64bits Rstudio Version 1.4.1106

Answer 1

Can use format to specify the display format:

n <- c(25351001641201357706367, 72952982679250725702754)

as.character(n)
#> [1] "2.53510016412014e+22" "7.29529826792507e+22"

as.character(format(n, scientific = FALSE))
#> [1] "25351001641201359126528" "72952982679250725240832"

Or use options to change the limit for scientific notation to a fairly high value, so that it is not used during the entire working session:

options(scipen = 1e6)

as.character(n)
#> [1] "25351001641201359126528" "72952982679250725240832"

Updated: large number of digits

Of aid to integer (free translation mine):

Note that current implementations of R use 32-bit integers for integer vectors. Therefore, the range of representable integers is restricted to about +/-2*10 9.

You can check the exact limit with .Machine$integer.max. Higher values are automatically converted to double:

n <- 25351001641201357

print(n, digits = 22)
#> [1] 25351001641201356

class(n)
#> [1] "numeric"

is.integer(n)
#> [1] FALSE

is.double(n)
#> [1] TRUE

Why this requires a very long explanation for an OS response, but the calculation of the number of significant digits in R depends on the implemented C library and follows international standard (can read more on Wikipedia in English).

The solution to this is to load the data as character. Because R automatically converts high values to double, even packages that implement higher limits (e.g. gmp) require them to be read first as a string.

Answer 2

options(scipen = 999)

df <- data.frame(Processo = (25351001641201357))
df