In fact, I have made long time ago the analysis of all the theorems and  tests of the Aslaksen paper for Maple. It intended to be a blog or similar, but did not had time yet to write it. As a sub product of this analysis I came across with the arctan/floor/ceil issues, which should have been yet another blog.  

In fact, I have made long time ago the analysis of all the theorems and  tests of the Aslaksen paper for Maple. It intended to be a blog or similar, but did not had time yet to write it. As a sub product of this analysis I came across with the arctan/floor/ceil issues, which should have been yet another blog.  

Your description sounds to me similar to this other problem. So, my advice would also be to try with (32 bit) Classic GUI (perhaps following this blog).

Probably, a good form for the output of evalc(argument(exp(y*I))) could be y-2*Pi*ceil((y-Pi)/(2*Pi)), plus an assumption that y is real:

ar2:=y->y-2*Pi*ceil((y-Pi)/(2*Pi)):
ar2~([Pi,-Pi,-Pi+.1,2*Pi,3*Pi]);
                      [Pi, Pi, -Pi + 0.1, 0, Pi]

Except for some quirks in ceil ...

Probably, a good form for the output of evalc(argument(exp(y*I))) could be y-2*Pi*ceil((y-Pi)/(2*Pi)), plus an assumption that y is real:

ar2:=y->y-2*Pi*ceil((y-Pi)/(2*Pi)):
ar2~([Pi,-Pi,-Pi+.1,2*Pi,3*Pi]);
                      [Pi, Pi, -Pi + 0.1, 0, Pi]

Except for some quirks in ceil ...

Certainly, that is the correct usage in this context. And quite frequently, by abuse of language, velocity is used when speed is meant.

Certainly, that is the correct usage in this context. And quite frequently, by abuse of language, velocity is used when speed is meant.

I am not surprised by your examples. Quite frequently, the output of convert is not an object computationally identical to the input object (or that represents the same mathematical object). And the combination of implicit and explicit assumptions is kind of patchwork.

Instead of deleting, you can select:

lprint(op(2,RTABLE(150112496,MATRIX([[1, 0, 0], [0, 1, 0], 
[0, 0, 1]]),Matrix)));

MATRIX([[1, 0, 0], [0, 1, 0], [0, 0, 1]])

Instead of deleting, you can select:

lprint(op(2,RTABLE(150112496,MATRIX([[1, 0, 0], [0, 1, 0], 
[0, 0, 1]]),Matrix)));

MATRIX([[1, 0, 0], [0, 1, 0], [0, 0, 1]])

I think that a representation for modulo a symbolic constant is missing.

I think that a representation for modulo a symbolic constant is missing.

So, you say that the output of evalc(argument(exp(y*I))) should be y with the property RealRange(Open(-Pi),Pi) ?

So, you say that the output of evalc(argument(exp(y*I))) should be y with the property RealRange(Open(-Pi),Pi) ?

Something seems missing. In which way you say  "using evalc
even produces values beyond Pi."?