Is it possible to Vectorize functions that output bigz objects?

user_10319707 0
Viewed 94

I suspect that there must be some way, for example gmp's factorialZ seems pre-vectorized:

> library(gmp)
> factorialZ(0:9)
Big Integer ('bigz') object of length 10:
 [1] 1      1      2      6      24     120    720    5040   40320  362880

and it appears to be comfortable being used with base R functions that take a vector input and give vector output

> cumsum(factorialZ(0:9))
Big Integer ('bigz') object of length 10:
 [1] 1      2      4      10     34     154    874    5914   46234  409114

However, probably because of coercion, attempting to Vectorize a function that outputs bigZ objects will be met with horrible failure:

leftFactorial<-function(n)
{
  sum(factorialZ(0:(n-1)))
}
> Vectorize(leftFactorial)(1:10)
  [1] 01 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 02 00 00 00 01 00 00
 [36] 00 01 00 00 00 01 00 00 00 04 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 0a 00 00 00 01 00 00 00 01 00
 [71] 00 00 01 00 00 00 22 00 00 00 01 00 00 00 01 00 00 00 01 00 00 00 9a 00 00 00 01 00 00 00 01 00 00 00 01
[106] 00 00 00 6a 03 00 00 01 00 00 00 01 00 00 00 01 00 00 00 1a 17 00 00 01 00 00 00 01 00 00 00 01 00 00 00
[141] 9a b4 00 00 01 00 00 00 01 00 00 00 01 00 00 00 1a 3e 06 00

So what should we do when we want to Vectorize a function that outputs bigZ objects?

gmp factorial vectorization r
1 Answers

dput allows to visualize the internal structure of a bigZ object : a vector of raw

> dput(factorialZ(0:9))
structure(as.raw(c(0x0a, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 
0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 
0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 
0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 
0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 
0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x18, 0x00, 
0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x78, 
0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 
0xd0, 0x02, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 
0x00, 0xb0, 0x13, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 0x00, 
0x00, 0x00, 0x80, 0x9d, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x01, 
0x00, 0x00, 0x00, 0x80, 0x89, 0x05, 0x00)), class = "bigz")

Using Vectorize with SIMPLIFY = TRUE results in a simplified vector of raw objects.

As noted in comments, SIMPLIFY = FALSE works because it keeps the original Big Integer type:

Vectorize(leftFactorial, SIMPLIFY = FALSE)(1:10)

[[1]]
Big Integer ('bigz') :
[1] 1

[[2]]
Big Integer ('bigz') :
[1] 2

[[3]]
Big Integer ('bigz') :
[1] 4

[[4]]
Big Integer ('bigz') :
[1] 10

[[5]]
Big Integer ('bigz') :
[1] 34

[[6]]
Big Integer ('bigz') :
[1] 154

[[7]]
Big Integer ('bigz') :
[1] 874

[[8]]
Big Integer ('bigz') :
[1] 5914

[[9]]
Big Integer ('bigz') :
[1] 46234

[[10]]
Big Integer ('bigz') :
[1] 409114
user_135133280
Related