@tomleslie 

You wrote " Generally speaking ...  I'm more interested in answers, rather than the "color" of answers! "
So am I but in some circumstances displaying data in a clever way really helps to get an idea of what happens.

Do you want an example:right, take the procedure HeatMap from package Statistics. It displays data in a matrix form with "cells" colored in some way, for instance color=[blue, white, red].
Very useful to have a quick picture of a correlation matrix for instance.
Now suppose that you want to superimpose the value of the correlation coefficient. Then black characters on white, light red or blue background are easily readable.
But black characters on red or blue are not and printing the correlation coefficient in white is a lot easier to read.

So "the "color" of answers!" is not always cosmetics, it's even common practice in data analysis.

 

@dharr 

Right, I didn't mentioned it but fillcolor works perfectly well.
But I guess it's because fillcolor operates on cells (such as  'background' in maplets for instance) and color on the content of the cell ( ~ 'foreground' in maplets).

@shakuntala 

Agree.

But Maple won't convert by itself your bvp into a ivp one.
I have no doubt that Maple will solve it with the classic RK4 method once you have done the transformation by hand.

@becejac 

Yes, the corresponding commands correspond to the yellow-highlighted red text.
The corresponding undirected graph cannot have more edges than the directed one (for if this later contaoins edges [a, b] and [b, a]; the former will only contain {a, b} if a < b or {b, a} otherwise).
For your graph, the undirected one has just one edge less than the directed one, meaning there is only pair (a, b) such that [a, b] and [b, a] are edges of the directed graph.
This pair is (344, 346) (the corresponding subgraph is provided)
 

Download Excel2Graph_ter.mw

@shakuntala 

From the help pages (search for "rk4"): "The RungeKutta command is a shortcut for calling the InitialValueProblemcommand with the method = rungekutta option."
Then all attempt to use RK4 (std shortcut for 4th order Runge-Kutta method) in Maple will generate an error with your Boundary Value Problem.

This is not a restriction from Maple: all classical RK methods are aimed to solve numerically only Initial Value Problems (see for instance: https://en.wikipedia.org/wiki/Runge–Kutta_methods).
You still have the possibility to use RK4 (for insnatce) for a Boundary Value Problem, if you use a Shooting Method (there was a recent question on Mapleprimes about this method)

Sorry tomleslie if I enable myself to answer in your behalf: tom used  method = bvp[middefer], a method, as its acronym suggests, designed to solve Boundary Value Problems

@max125 

In case you would like to see all the digits type lprint(A)

@acer 

If E is an ellipsoid of equation x^2 + y^2/4+z^2/9 = k (k >= 0) the solution is even easier to obtain:

 

Download MultivariateGaussian.mw

PS: for info, this result is at the core of  FORM (First Order Reliabiliy Method) widely used in ingeneering science to assess the probability of failure of a system.

PS@PS: I wrote this worksheet while looking the soccer champion's league:I hope I didn't do any mistake

@Carl Love 

Sorry, typed in a hurry, "If you want to integrate over a box of R^3, the result is also obvious."
where a box is defined here the set of all (x, y, z) in R^3 such that a <= x <= b,  a' <= y <=b', a'' <= z <= b'' (I'm not sure that the name "box" ihas everywhere the same meaning

@Carl Love 

Thanks.

In fact I could formulate the same answer I just did to acer: it's difficult fo find informations in the help pages about unicodes and how to use them.
For instance, how did you infer that 20ac was to be precede by &#x ?

@acer 

I will test your file tomorrow with Maple 2018... but the  graphics already seem very pretty.

The point is, you do extensive use of typeseting, mrow, mo like in typeset(seq([nprintf(`#mrow(mo("%a"),mo("&rarr;"));`, idvars[i]), "\n"][]
and it's hard finding information about this in the help pages.

@acer 

Thank you for the answer

@acer 


Nice to you to ask this question.

Unfortunately I can't load the example (which runs with Maple 2018) right now for I'm home.

So I prepared this illustrative example.
The last pink line at the end of the worksheet is related to the  "[a] graph (not the one GraphTheory:-Draw could produce, but this same graph after some modification based on ImageTools to display non ascii characters" part of my initial question.
Indeed, I found in ImageTools the non ascii characters I would have like to display in the variable "IDtext" .
So I thought it could be good idea to construct IDtext by using ImageTools:-text , to change POLYGONS into ImageTools:-Poly, and so on, to obtain at the end a "graph" not defined in terms of PLOT[structure] elemnts, but in terms of ImageTools elements. 
I'm aware that I may have gone to pick up noon to 2:00 pm, but that's what amateurs do.

Once the probabilistic model is sampled, a GridPlot object (Diagonal=Histogram, Upper=ScatterPlot, Lower=Correlation PieChart) and a Matrix wich summarizes the info about all the random variables names, laws, parameters, tructations, ...)  are  added to the graph below.
This is this whole collection of objets that I tried to display from a single procedure, in order to provide an exhaustive résumé of the whole probabilistic model... Wow, it's over


 

Download NAC.mw

@acer 

Hi, 

what do you submit a bug report?
It's not rare I meet behaviours that seem to look like this one. In general I put this on account of my limited knowledge of Maple and I try to act otherwise.
For example, in the situation you describe, (simplify@expand)(ee) returns the correct result and it would have been fine with me without me questioning a possible weakness in simplicity.
Sometimes people here give you advices this kind and there is (AFAIK) no decision of submission of a bug report?
So why, in this particular circumstance, you consider this worth it?

@mrakgr 

nm already gave you the answer,  here is, for instance,  the way you could write procedure f:

Download f.mw

Maybe you could try to do numerical integration with different methods, compare times and select the good one.
Using an external library will solve nothing if you do not use  the "correct" integration algorithm.
And this is always problem-dependent.

PS: for multiple integration Maple says it explicitely uses the "external" CUBA library.