## CRAN does not validate R packages!

Posted in pictures, R, University life with tags , , , , , , , , , , on July 10, 2019 by xi'an

A friend called me the other day for advice on how to submit an R package to CRAN along with a proof his method was mathematically sound. I replied with some items of advice taken from my (limited) experience with submitting packages. And with the remark that CRAN would not validate the mathematical contents of the associated package manual. Nor even the validity of the R code towards delivering the right outcome as stated in the manual. This shocked him quite seriously as he thought having a package accepted by CRAN was a stamp of validation of both the method and the R code. It would be nice of course but would require so much manpower that it seems unrealistic. Some middle ground is to aim at a journal or a peer community validation where both code and methods are vetted. Which happens for instance with the Journal of Computational and Graphical Statistics. Or the Journal of Statistical Software (which should revise its instructions to authors that states “The majority of software published in JSS is written in S, MATLAB, SAS/IML, C++, or Java”. S, really?!)

As for the validity of the latest release of R (currently R-3.6.1 which came out on 2019-07-05, named Action of the Toes!), I figure the bazillion R programs currently running should be able to detect any defect pretty fast, although awareness of the incredible failure of sample() reported in an earlier post took a while to appear.

## Le Monde puzzle [#1105]

Posted in Kids, R with tags , , , , , , on July 8, 2019 by xi'an

Another token game as Le Monde mathematical puzzle:

Archibald and Beatrix play with a pile of n>100 tokens, sequentially picking m tokens from the pile with m being a prime number [including m=1] or a multiple of 6, the winner taking the last tokens. If Beatrix knows n and proposes to Archibald to start, what is the value of n?

Which cannot be solved in a few lines of R code:

k<-function(n)n<4||all(n%%2:ceiling(sqrt(n))!=0)||!n%%6
g=(1:3)
n=c(4,i<-4)
while(max(n)<101){
if(k(i)) g=c(g,i) else{
while(i%in%g)i=i+1;j=4;o=!j
while(!o&(j<i)){
o=(j%in%n)&k(i-j);j=j+1}
if(o) g=c(g,i) else n=c(n,i)}
i=i+1}


since it returned no unsuccessful value above 100! With 4, 8, 85, 95, and 99 as predecessors. A rather surprising outcome and a big gap that most certainly has a straightforward explanation! Or a lack of understanding from yours truly: this post appears after the solution was published in Le Monde and I am more bemused than ever since the losing numbers in the journal are given as 4, 8, 85, … 89, and 129. With the slight hiccup that 89 is a prime number…. The other argument in the solution that there can only be five such losers is well-taken since there are only five possible non-zero remainders in the division by 6.

## Le Monde puzzle [#1104]

Posted in Kids, R with tags , , , , on June 18, 2019 by xi'an

A palindromic Le Monde mathematical puzzle:

In a monetary system where all palindromic amounts between 1 and 10⁸ have a coin, find the numbers less than 10³ that cannot be paid with less than three coins. Find if 20,191,104 can be paid with two coins. Similarly, find if 11,042,019 can be paid with two or three coins.

Which can be solved in a few lines of R code:

coin=sort(c(1:9,(1:9)*11,outer(1:9*101,(0:9)*10,"+")))
amounz=sort(unique(c(coin,as.vector(outer(coin,coin,"+")))))
amounz=amounz[amounz<1e3]


and produces 9 amounts that cannot be paid with one or two coins.

21 32 43 54 65 76 87 98 201

It is also easy to check that three coins are enough to cover all amounts below 10³. For the second question, starting with n¹=20,188,102,  a simple downward search of palindromic pairs (n¹,n²) such that n¹+n²=20,188,102 led to n¹=16,755,761 and n²=3,435,343. And starting with 11,033,011, the same search does not produce any solution, while there are three coins such that n¹+n²+n³=11,042,019, for instance n¹=11,022,011, n²=20,002, and n³=6.

## another attempt at code golf

Posted in Books, Kids, R with tags , , , on June 12, 2019 by xi'an

I had another lazy weekend go at code golf, trying to code in the most condensed way the following task. Provided with a square matrix A of positive integers, keep iterating the steps

• take the highest square $$𝑥²$$ in A.
• find the smallest adjacent neighbour $$𝑛$$
• replace with x and n with nx

until no square is left (with neighbour defined as either horizontally or vertically and without wrapping around). While I managed a 217 bytes solution, compared with Robin’s 179b improvement, which remains surprising readable!, the puzzle offers two further questions:

1. is there a non-iterative way to find the final matrix B?
2. the puzzle assumes that A satisfies that at each step, the highest square and the smallest neighbour n will be unique, and that the sequence will not repeat forever. Is there a fool-proof way to check this is the case?

## biased sample!

Posted in Statistics with tags , , , , , , , , , , , on May 21, 2019 by xi'an

A chance occurrence led me to this thread on R-devel about R sample function generating a bias by taking the integer part of the continuous uniform generator… And then to the note by Kellie Ottoboni and Philip Stark analysing the reason, namely the fact that R uniform [0,1) pseudo-random generator is not perfectly continuously uniform but discrete, by the nature of numbers on a computer. Knuth (1997) showed that in this case the range of probabilities is larger than (1,1), the largest range being (1,1.03). As noted in the note, exploiting directly the pseudo-random bits of the pseudo-random generator. Shocking, isn’t it!  A fast and bias-free alternative suggested by Lemire is available as dqsample::sample

As an update of June 2019, sample is now fixed.

## an attempt at code golf

Posted in Kids, R with tags , , , , , , , , , on May 15, 2019 by xi'an

Having discovered codegolf on Stack Exchange a few weeks ago, I spotted a few interesting puzzles since then but only got the opportunity at a try over a quiet and rainy weekend (and Robin being on vacation)! The challenge was to write an R code for deciding whether or not a given integer n is congruent or not, when congruent means that it is the surface of a rectangle triangle with all three sides rational. The question included a pointer to the Birch and Swinnerton-Dyer conjecture as a mean to check congruence although the real solution was provided by Tunnell’s Theorem, which states that n is congruent if and only if the number of integer solutions to 2x²+y²+8z²=n is twice as much as the number of integer solutions to 2x²+y²+32z²=n if n is odd and  the number of integer solutions to 8x²+y²+16z²=n is twice as much as the number of integer solutions to 8x²+y²+64z²=n if n is even. Although this is only true for squared-free integers. (I actually spent more time on figuring out the exact wording of the theorem than on optimising the R code!)

My original solution

p=function(n){
for (i in(n:2)^2)if(n%%i<1)n=n/i
if(n%%2){d=8;f=2;g=16}else{d=2;f=1;g=8}
A=0;b=(-n:n)^2
for(x in d*b)for(y in x+f*b)for(z in g*b)
A=A+(y+z==n)-2*(y+4*z==n)
A==0}


was quite naïve, as shown by the subsequent improvements by senior players, like the final (?) version of Guiseppe:

function(n){b=(-n:n)^2
for(i in b[b>0])n=n/i^(!n%%i)
P=2^(n%%2)
o=outer
!sum(!o(y<-o(8/P*b,2*b,"+")/P-n,z<-16/P*b,"+"),-2*!o(y,4*z,"+"))}


exhibiting a load of code golf tricks, from using an anonymous function to renaming functions with a single letter, to switching from integers to booleans and back with the exclamation mark.

## chance call for book reviewers

Posted in Statistics with tags , , , , , , , , on May 14, 2019 by xi'an

Since I have been unable to find local reviewers for my CHANCE review column of the above recent CRC Press books, namely