I have been writing a language translator. I have everything working reasonably well except that when translating abs(expr) with optimize=tryhard, the IntermediateCode converts the abs() into if...then...else statements. This would be fine except the type of the expr is generally a complex number. I need to optimize the code because there is a lot of redundant calculations otherwise.

I can't figure out how to get IntermediateCode from breaking the abs function into if...then...else statements. [My solution so far is to substitute abs with a dummy name and then use the translator to translate the dummy name into an abs statement - That's really not how things should work!]

Any suggestions?

I've included some example test cases of what is going wrong.

Download intermediate_code_abs.mw

Hi,

I'm trying to figure out how to get the HermitianTranpose vector to distribute over a sum of vectors. Maple seems to want to ignore the linearity of the conjugate operation.

See the example code attached.

Thanks.
 

Download DistributeHermitian.mw

By default, CodeGeneration generates temporary names during optimize of t####. For compatibility with other code, I would like to have flexibility in choosing the preamble to the temporary variable name ("t" is default). Is there a way to do this?

I don't see anything in the documentation, but it seems like something one should be able to do.

(My target language is Python - although I'm not really using Python, the syntax is compatible).

I am trying to get Maple to recognize that

diff(x^n,x) does not equal n*x^(n-1), but rather 0 if n=0, or n*x^(n-1) otherwise.

This comes up when differentiating an infinite sum (power series). The constant term gets transformed into n/x instead of becoming zero. Maybe this is really a bug/lack of feature in how sum/Sum works.

For example: diff(Sum(x^n, n = 0 .. infinity), x) yields

Does anybody have a fix to get the differentiation right (other than expanding some terms of the series before taking the derivative)? 

(edited) Please note I am not trying to get an answer to this specific question. It is just illustrates a simplified example of behavior that leads to other bizarre results. Please see the attached worksheet for more weirdness: Problems_with_summation_Mapleprimes.mw

Hi,

I just upgraded to Maple 2021 and started getting strange results - similar calculations seemed to work in Maple 2020.

I am looking at correlated bivariate normal distributions. When doing a fully symbolic verification of the normalization, I get an incorrect result of infinity. I also get strange results when integrating over a circular region, but the normalization seemed to be a very elementary calculation.

Maybe someone can check this make sure I'm not losing my marbles. I've attached a worksheet.

Thanks.

Bivariate_Gaussian.mw