You have one equation with 14 variables. If you can't come up with some numeric values for these so-called "constants", then the situation is hopeless.

@want to be a permanent vegan 

You say that "it" doesn't work. I want to make it clear that "it" refers to your functions, not my solution to your previous problem. My solution does indeed work in the sense that it corrects the error that you were getting before, which was a Maple error. This new error is a deeper more-mathematical error in your functions, which I'm too tired to delve into right now.

One way of achieving the piecewise plot that you want is this:

plots:-display([plot(I1, 0..100), plot(I2, 100..200)]);

Once again, the above will work if and only if your functions can be evaluated to real values in their respective intervals, which I haven't checked.

@Markiyan Hirnyk 

I was a member of one of the perfect-score teams in the contest, and the only member of that team who was expert in Maple. I completed four of the problems in Maple, one of which was the problem we're discussing now.

All of the problems were, by design, probably impossible to do directly with any CAS in existence at the time. I'm not sure that Prof. Lloyd N. Trefethen (the one who posed the problems) actually checked every CAS, but he was (is) a renown expert in such matters. By "doing the problem" I mean, for all ten problems, obtaining ten decimal digits of approximation; symbolic solutions are not wanted. All of the problems required (at that time) a significant amount of computation and a significant amount (many hours or days) of "by hand" work to get the ten digits. Much faster solutions have subsequently been found for most of the problems.

It's possible that Mathematica has made good use of the wealth of knowledge provided by the many published solutions to the problems (a whole book has been written about these solutions); I believe that Maple has failed to make use of that knowledge. The book is The SIAM 100-Digit Challenge: A Study in High-Accuracy Numerical Computing by Folkmar Bornemann, Dirk Laurie, Stan Wagon, and Jorg Waldvogel; 2004 SIAM (the Society for Industrial and Applied Mathematics); ISBN 0-89871-561-X.

@Markiyan Hirnyk Yes, if a Maple table is truly desired, it would be done like that. I was just guessing that the OP was using table colloquially, not in the Maple sense of the word. But you may be right.

@Markiyan Hirnyk Nice use of DirectSearch to find complex roots of a complex-valued function. I converted your Reply into an Answer so that I could vote it up.

Regarding your inability to upload the file to MaplePrimes: MaplePrimes has a problem with certain special characters in filenames.

Regarding my Answer: My main purpose was to show the OP how to improve his code that filtered the roots given by fsolve. I'm not at all surprised that fsolve missed many roots. I am suprised by the abyssmal failure of RootFinding:-Analytic to find any roots at all.

@Markiyan Hirnyk With fsolve, I found 16 roots in just the third quadrant of the OP's rectangle. These are shown in my Answer below. My first attempt was also to use RootFinding:-Analytic. I tried many different ways, but couldn't find a single root with it. I tried several smaller regions over which the function is indeed analytic.

Regarding your plot, I believe that the function is too "wavy" for its behavior to be adequately captured on the coarse grid of your plot. The magnitude has many very steep and narrow "valleys", as shown on my plots in my Answer.

@acer I don't know if this is a 2D-input difference, but don't you mean Shift-Enter rather than Ctrl-Enter? For me (using Maple Input (1D)), Ctrl-Enter inserts a page break.

I appreciate that you uploaded your worksheet for us to read. It'd be helpful if you'd also attach it to your post so that we could download it to work with it in our own Maple.

@sunflower 

1. Don't remove a Reply after it has already been replied to. That makes my previous Reply seem out of context. You had asked for an explanation of the lines where I defined rectangle.

2. If you want another house, you'll need to tell me what to put on it. I don't have much artistic imagination. And why do you want another house? How about just asking a question directly?

3. I have no idea what you mean by "form".

4. Your English doesn't make me unhappy.

@sunflower I don't know if this'll be much help to you, but those two lines mean the same thing as

rectangle:= proc(Point, height, width)
     plots:-polygonplot(
          [Point, Point+[width,0], Point+[width,height], Point+[0,height], Point],
          args[4..]
     )
end proc:

You should look up ?parameter in the Maple help. That should keep you occupied for a few days.

The integral I5 clearly converges for any real x0! I'd expect even my good calculus II students to see that (without using Maple). The expression under the radical is strictly positive, so there are no poles to worry about. Do a limit comparison test with Int(1/x^4, x= 1..infinity).

P.S.: I just saw that Preben said essentially the same thing.

Your example doesn't make sense. What does Oy mean? What does "limit by y, x=1, y=5" mean?

To understand what a tubeplot is, you must first understand what a spacecurve is. Look up the help for spacecurve (?spacecurve). Let me know if you understand that. If you don't understand that, we'll need to start there.

@tomleslie Tom's Answer is the true answer (an exposition of The Way), not a fake. If you want to do symbolic computation, don't use floats. His Answer works on irrational bases also, despite what he says, and any float can be converted to a rational.

Kitonum's Answer uses a fake technique. This is not meant as a disparagement of Kitonum's Answer---the fake can be useful sometimes.

@Markiyan Hirnyk I think that four digits accuracy is pretty darn good from a plot which consists of exactly 225 points, that is, only three digits worth of points. It is big win for plot that it found this small turn in the function. Another way of looking at it is that your maximum and my maximum are visually indistinguishable on the scale of the plot. And why should you expect a plot to contain information that can't be visually distinguished? My program is intended to find the extremal points that actually occur on plots, not the extrema of analytic functions. You intentionally poison the well with your intentionally irrelevant comparisons.

@Markiyan Hirnyk I made a small adjustment to my plaintext code. I believe that now when you paste it into 2D input it'll work (although I have no interest in testing this myself). Give it another try. I do always expect/hope for my plaintext code to work in 2D input, although I personally wouldn't touch the 2D input with a ten-foot pole for programming. For static display of mathematical expressions, it's okay, not great.