I guess it might be clearer by now, that "normal distribution of points around a line" is ambiguous. What exactly is the random variable(s) that you intend? Is it a distance of separation normal from the curve, or some purely y-separation, or purely x-separation, or the sum of both, or some other bivariate RV?

Don't both of those example have a flaw in that they only add noise is one direction? If I read it right, they add a real component of noise, but omit to add an imaginary component of noise. If that's right, then the scatter plot from them would look flat and non-noisy along two opposite sides, and only fuzzy in between those sides.

Also, the noise added in the first example is perhaps too large in scale, and swamps the data signal?

Since complexplot is just going to extract the real and imaginary components as it produces the final plot, then it seems a roundabout way of doing things to build cos(k*t) and I*sin(k*t) when the two purely real x- and y-components could just as easily be put directly and right away into the desired final form that plot can handle (with each getting its own noise component0.

It's usually better from a performance viewpoint to generate whole Vectors as a sample from a distribution, up front, and then index into it as required, rather than to generate single-element samples one at a time and repeatedly.

Don't both of those example have a flaw in that they only add noise is one direction? If I read it right, they add a real component of noise, but omit to add an imaginary component of noise. If that's right, then the scatter plot from them would look flat and non-noisy along two opposite sides, and only fuzzy in between those sides.

Also, the noise added in the first example is perhaps too large in scale, and swamps the data signal?

Since complexplot is just going to extract the real and imaginary components as it produces the final plot, then it seems a roundabout way of doing things to build cos(k*t) and I*sin(k*t) when the two purely real x- and y-components could just as easily be put directly and right away into the desired final form that plot can handle (with each getting its own noise component0.

It's usually better from a performance viewpoint to generate whole Vectors as a sample from a distribution, up front, and then index into it as required, rather than to generate single-element samples one at a time and repeatedly.

Hard to say much more, with the full source of your example. What is Digits set to?

You could try EvalhfMapleProc instead, to make the callback be into evalhf.

Hard to say much more, with the full source of your example. What is Digits set to?

You could try EvalhfMapleProc instead, to make the callback be into evalhf.

@Alejandro Jakubi We've discussed at least two bugs here.

One is that in Maple 15 the double-quotes must be escaped when calling maplew.exe to open bar.mpl on the commandline, by making the filename the only argument. When I wrote of a regression, I meant this bug, as it worked properly without escaping the double-quotes in Maple 14.

Another bug discussed is the need for extra, escaped double-quotes when using the -c option to make a `read` statement.

A related bug is that the Standard GUI will not open a .mpl file from the usual MS-Windows file-manager. And another bug (also a regression as it worked in previous versions) is that the Standard GUI does not open .mla files in Maple 15/15.01 when using the usual association from the usual file-manager (but File->Open from the Maple menu bar works, invoking march/open as appropriate).

@Alejandro Jakubi We've discussed at least two bugs here.

One is that in Maple 15 the double-quotes must be escaped when calling maplew.exe to open bar.mpl on the commandline, by making the filename the only argument. When I wrote of a regression, I meant this bug, as it worked properly without escaping the double-quotes in Maple 14.

Another bug discussed is the need for extra, escaped double-quotes when using the -c option to make a `read` statement.

A related bug is that the Standard GUI will not open a .mpl file from the usual MS-Windows file-manager. And another bug (also a regression as it worked in previous versions) is that the Standard GUI does not open .mla files in Maple 15/15.01 when using the usual association from the usual file-manager (but File->Open from the Maple menu bar works, invoking march/open as appropriate).

It's really not good to build lists like this, repeatedly appending inside a loop, if you expect to generate a final result with many entries. It will just be much slower than it has to be, and use more memory than it has to use, and it won't scale well to a very large number of points.

It's really not good to build lists like this, repeatedly appending inside a loop, if you expect to generate a final result with many entries. It will just be much slower than it has to be, and use more memory than it has to use, and it won't scale well to a very large number of points.

Thank you, Joe.

I don't see why the 'symbolic' option should be necessary for this example. Maybe it shouldn't be.

(Sometimes it's useful to get such a specific effect as combine/radical offers, without the broader effects and cost of calling `simplify`.)

Thank you, Joe.

I don't see why the 'symbolic' option should be necessary for this example. Maybe it shouldn't be.

(Sometimes it's useful to get such a specific effect as combine/radical offers, without the broader effects and cost of calling `simplify`.)

Thank you, Preben. And apologies to the forum, for wasting people's time.

I was looking at a larger expression from this problem, which I thought had such terms and which I thought was not simplifying with just `simplify` alone. I obviously wasn't thorough in my checking, when I tried the smaller example!

Thank you, Preben. And apologies to the forum, for wasting people's time.

I was looking at a larger expression from this problem, which I thought had such terms and which I thought was not simplifying with just `simplify` alone. I obviously wasn't thorough in my checking, when I tried the smaller example!

@PatrickT That site is reposting material from some other (legitimate) math forum, that was originally posted about 2006, but the posts are tagged as being from the period 2010-2011.

I would say that it is disingenuous for any site to be making it look like such old material is newly from 2010-2011.

That site is very heavily loaded with commercial advertisements. I wouldn't be surprised if advertising were not the site's principal purpose.

I wonder whether there is a trademark issue, with that site's name.

@PatrickT That site is reposting material from some other (legitimate) math forum, that was originally posted about 2006, but the posts are tagged as being from the period 2010-2011.

I would say that it is disingenuous for any site to be making it look like such old material is newly from 2010-2011.

That site is very heavily loaded with commercial advertisements. I wouldn't be surprised if advertising were not the site's principal purpose.

I wonder whether there is a trademark issue, with that site's name.