5 ms·

I'm always hearing about these and other problems that programmers have with the R language. In practice, none of them seem to be a big deal to statisticians.

I'm always hearing about these and other problems that programmers have with the R language. In practice, none of them seem to be a big deal to statisticians. 64 bit integers? Never had a need for this, but maybe there is a package that can add them in? What's wrong with the package system? I maintain an R package on CRAN and I'm quite smitten with the package system.

z2210558 8y ago
I think bit64 gives integer64, but like you I never need large integers - I don't do general purpose programming in R.
- infogulch 8y ago
  What is the overflow behavior? If it fails and warns you that some operation overflowed, I don't see an issue.
  - disgruntledphd2 8y ago
    It converts really large integers to floats, which can lose precision. You need to handle really large numbers as characters, or use bit64. I only really started hitting this in the private sector though, never a problem when I was in academia.
    - infogulch 8y ago
      Oh. That's kindof... gross. Better than wrapping silently I guess. Can you at least tell when a number has been 'promoted' to a float so you know you need a bigger type?
another-one-off 8y ago
Also fairly critically; a mathematician can go from assuming that one exists as a concept to multiple ordinals of infinity with about 20 minutes of scribbling. The fact that 64 bit is a modern standard doesn't exclude the fact that R cannot under any circumstance handle the vast and overwhelming majority of known numbers. I feel sorry for the poor souls who turn out to need it, though. R sucks like that.
bsg75 8y ago
> 64 bit integers? Never had a need for this, but maybe there is a package that can add them in? https://cran.r-project.org/web/packages/bit64/index.html https://cran.r-project.org/web/packages/bit64/index.html
srean 8y ago
> I'm always hearing about these and other problems that > programmers have with the R language. In practice, none > of them seem to be a big deal to statisticians. > 64 bit integers? Never had a need for this This kind of thinking is pervasive in a wide corner of statistics and this is pretty much why machine learning stole its crown, that's why I get to hear quips like "Statistics ? Is that even relevant ?"
- nightski 8y ago
  Crown of what? They are complementary fields. Try doing ML without any statistics and lets see how far you get.
  - srean 8y ago
    They are closer than complementary and that is exactly what I am trying to point out. How much funding, support, mind share, conferences, venture capital yadda yadda does stats get compared to ML -- thats what I mean by losing its crown. All that ML has now could easily have been stat's were it not for the attitude I draw attention to. Isnt that sad, given that ML in many cases is just stats without the baggage "R can solve all of my problems and whatever R cant solve isnt a stats problem"
    - nightski 8y ago
      Yes I agree wholeheartedly, sorry I originally missed your point. I do think it's more of a DL vs. the world problem than stats vs. ML but everything you said rings true. I was kind of odd in that I started with DL and then became fascinated with statistics and lost focus with DL. Statistics is just so much more useful outside of the niche domains where DL is extremely effective.
      - srean 8y ago
        > Statistics is just so much more useful outside of the niche domains where DL is extremely effective. This. Sorry had to.

z2210558 8y ago

I think bit64 gives integer64, but like you I never need large integers - I don't do general purpose programming in R.

infogulch 8y ago
What is the overflow behavior? If it fails and warns you that some operation overflowed, I don't see an issue.
- disgruntledphd2 8y ago
  It converts really large integers to floats, which can lose precision. You need to handle really large numbers as characters, or use bit64. I only really started hitting this in the private sector though, never a problem when I was in academia.
  - infogulch 8y ago
    Oh. That's kindof... gross. Better than wrapping silently I guess. Can you at least tell when a number has been 'promoted' to a float so you know you need a bigger type?

another-one-off 8y ago

Also fairly critically; a mathematician can go from assuming that one exists as a concept to multiple ordinals of infinity with about 20 minutes of scribbling. The fact that 64 bit is a modern standard doesn't exclude the fact that R cannot under any circumstance handle the vast and overwhelming majority of known numbers. I feel sorry for the poor souls who turn out to need it, though. R sucks like that.

bsg75 8y ago

> 64 bit integers? Never had a need for this, but maybe there is a package that can add them in? https://cran.r-project.org/web/packages/bit64/index.html https://cran.r-project.org/web/packages/bit64/index.html

srean 8y ago

> I'm always hearing about these and other problems that > programmers have with the R language. In practice, none > of them seem to be a big deal to statisticians. > 64 bit integers? Never had a need for this This kind of thinking is pervasive in a wide corner of statistics and this is pretty much why machine learning stole its crown, that's why I get to hear quips like "Statistics ? Is that even relevant ?"

nightski 8y ago
Crown of what? They are complementary fields. Try doing ML without any statistics and lets see how far you get.
- srean 8y ago
  They are closer than complementary and that is exactly what I am trying to point out. How much funding, support, mind share, conferences, venture capital yadda yadda does stats get compared to ML -- thats what I mean by losing its crown. All that ML has now could easily have been stat's were it not for the attitude I draw attention to. Isnt that sad, given that ML in many cases is just stats without the baggage "R can solve all of my problems and whatever R cant solve isnt a stats problem"
  - nightski 8y ago
    Yes I agree wholeheartedly, sorry I originally missed your point. I do think it's more of a DL vs. the world problem than stats vs. ML but everything you said rings true. I was kind of odd in that I started with DL and then became fascinated with statistics and lost focus with DL. Statistics is just so much more useful outside of the niche domains where DL is extremely effective.
    - srean 8y ago
      > Statistics is just so much more useful outside of the niche domains where DL is extremely effective. This. Sorry had to.