## 432 Reputation

15 years, 154 days

## ErrorPlot...

It seems that I successfully embarrassed myself here. Not sure why I was blinded. Apologies to one and all.

## On assuming...

`limit(sin(k*Pi), k=infinity) assuming k::integer` does not work. (This is contrary to your comment.) `limit((sin(k*Pi) assuming k::integer), k=infinity)` does work. (Presumably this is because `sin(k*Pi) assuming k::integer` gets simplified to `0`.) `limit(sin(k*Pi) assuming k::integer, k=infinity)` does not work. (Presumably the `k=infinity` is passed to `assuming`.) The second and third behaviors make sense. The first behavior is less than ideal, but might just mean that `limit` ignores assumptions, like `solve`.

## On assuming...

`limit(sin(k*Pi), k=infinity) assuming k::integer` does not work. (This is contrary to your comment.) `limit((sin(k*Pi) assuming k::integer), k=infinity)` does work. (Presumably this is because `sin(k*Pi) assuming k::integer` gets simplified to `0`.) `limit(sin(k*Pi) assuming k::integer, k=infinity)` does not work. (Presumably the `k=infinity` is passed to `assuming`.) The second and third behaviors make sense. The first behavior is less than ideal, but might just mean that `limit` ignores assumptions, like `solve`.

## limit assumptions...

I fully agree. ((I did not realize that the extra parentheses were necessary.))

## limit assumptions...

I fully agree. ((I did not realize that the extra parentheses were necessary.))

## limit assumptions...

For me, the syntax of `> limit( (sin(k*Pi) assuming k::integer), k=infinity );` is fine. I think that this should be explicitly documented in the Help system though; an example in the Help for `limit` and `assuming` would be enough. Also, it seems to only work in 1D input, not 2D—a bug, I assume.

## limit assumptions...

For me, the syntax of `> limit( (sin(k*Pi) assuming k::integer), k=infinity );` is fine. I think that this should be explicitly documented in the Help system though; an example in the Help for `limit` and `assuming` would be enough. Also, it seems to only work in 1D input, not 2D—a bug, I assume.

## different types of irreproducibility...

There are at least three different types of irreproducibility (in addition to those associated with RandomTools). One type derives from mathematics. An example is {a,b,c}: sets are unordered, and so we should not expect reproducibility. Irreproducibility follows from some mathematical definitions. So if you rely on reproducibility here, the error is in your programming (as Robert Israel is indicating). A second type derives from using computers. With floating-point numbers, a+(b+c) is not always the same as (a+b)+c. So if SF is a set of floats, then add(i,i=SF) will give different results depending on the order in which the elements of SF are added. That order is inherently undefined; so the add is inherently irreproducible, in principle. There is also a third type, which derives from bugs. As an example (found by V. Bondarenko), assign > xpr:= 1/(sqrt(1+z)*sqrt(1+z^5)): and then execute > fnormal(evalf(int(xpr, z = 0..infinity), 20)); several times. Maple (11.01) gives different results, but it should not.

## different types of irreproducibility...

There are at least three different types of irreproducibility (in addition to those associated with RandomTools). One type derives from mathematics. An example is {a,b,c}: sets are unordered, and so we should not expect reproducibility. Irreproducibility follows from some mathematical definitions. So if you rely on reproducibility here, the error is in your programming (as Robert Israel is indicating). A second type derives from using computers. With floating-point numbers, a+(b+c) is not always the same as (a+b)+c. So if SF is a set of floats, then add(i,i=SF) will give different results depending on the order in which the elements of SF are added. That order is inherently undefined; so the add is inherently irreproducible, in principle. There is also a third type, which derives from bugs. As an example (found by V. Bondarenko), assign > xpr:= 1/(sqrt(1+z)*sqrt(1+z^5)): and then execute > fnormal(evalf(int(xpr, z = 0..infinity), 20)); several times. Maple (11.01) gives different results, but it should not.

## default input format...

I agree with Joe Riel.

## default input format...

I agree with Joe Riel.

## other languages...

That's a good point about people who can read English, but do not feel comfortable writing it. I have been in similar situations. Maybe I got a little sensitive because in Canada some people try to push French for political reasons. (The claim that "plenty of people can read French" is unrealistic if the reference set is Maple users.) Also, you say that there is no problem, because of BabelFish et al. Have you tried BabelFish translations of Japanese (your example)? I cannot understand such translations. So there is a problem. But I still agree with your main point.

## other languages...

That's a good point about people who can read English, but do not feel comfortable writing it. I have been in similar situations. Maybe I got a little sensitive because in Canada some people try to push French for political reasons. (The claim that "plenty of people can read French" is unrealistic if the reference set is Maple users.) Also, you say that there is no problem, because of BabelFish et al. Have you tried BabelFish translations of Japanese (your example)? I cannot understand such translations. So there is a problem. But I still agree with your main point.

## language...

인천 남구 용현동에 위치한 4년제 사립대학으로써, 1954년 인하공과대학으로 설립되어, 현재 공과대학·자연과학대학·경상대학·경영대학·사범대학·법과대학·사회과학 If postings are accepted in French, then how about other languages, such as Korean? My point is that it is not feasible for MaplePrimes readers to know the native languages of all Maple users. The only natural language that MaplePrimes readers can be assumed to know is English. Note 1: English is the language of international scientific discourse (I did not choose this, and would not have chosen English if I had the choice). Note 2: I have no idea what the Korean quote says.

## language...

인천 남구 용현동에 위치한 4년제 사립대학으로써, 1954년 인하공과대학으로 설립되어, 현재 공과대학·자연과학대학·경상대학·경영대학·사범대학·법과대학·사회과학 If postings are accepted in French, then how about other languages, such as Korean? My point is that it is not feasible for MaplePrimes readers to know the native languages of all Maple users. The only natural language that MaplePrimes readers can be assumed to know is English. Note 1: English is the language of international scientific discourse (I did not choose this, and would not have chosen English if I had the choice). Note 2: I have no idea what the Korean quote says.
﻿