So if a solution can eventually be found shouldn't Maple exhaust all methods behind the scenes to give a solution?  That is probably the holy grail coding of any CAS system

@Kitonum thanks for a solution

@acer I see.  I didn't come across anything about him mentioning that the fails didn't mean the CAS couldn't actually solve the problem.  I falsely undertook his findings as hard and true (probably in most cases other people as well). 

I think it basically boils down to not being able to do this

int(1/(sqrt(x+exp(x))),x)

I tried with subs which partially worked.  But as soon as I invoked the sort command the order jumbled out of order, until I used the restart command which makes my idea somewhat useless.

@Axel Vogt indeed, a one function approximation is probably enough, but as you say it looks more like 2 or even 3 functions.  I wonder if the op could let us in on what the values are supposed to represent.

On top of a wave increasing in amplitude it does appear the there is some underlying fluctuating superimposed wave.

I don't know, but I wonder if windspeed data could be pulled from https://earth.nullschool.net

Two thumbs up for the clustering and geographic additions!  This new edition looks like it will be a joy to explore.  I've been lost for innovative ideas in Maple the last couple of years, I hope to apply some ideas in the new Maple and to compile some things to apply and to improve on.

@Bryon thank you

It seemed like an interesting enough question - homework regardless - it does appear the question was mistakenly identified as spam.

Can the administrators repair?

As a relatively avid maple user, some time ago I required the use of Geometry expressions, to find an equation based on various angles and lengths with respect to something else.  I could have done it by hand, but of course it was easier to draw the pictures in Geometry expressions.  This could not be done so easily with Maple I recall. 

Can you present a screenshot?  Also provide Maple version and OS.  Probably just a restart will suffice.

It is something that is impossible to occur.

@ Kitonum - It does not work in Maple 12.

Error, (in pdsolve/sys) unable to handle PDE problem subject to boundary conditions {u(0, y) = 0, (D[1](u))(0, y) = 0}

However it does work in Maple 18.  Therefore this is a regression bug that exists in 2016.1  No one has been able to confirm if it exists in 2016.0 or not.  I will be able to confirm this tomorrow. 

@farjad_etg it's no different to what you originally had. 

So what we've concluded is 2016.1 won't solve the it using pdsolve but will using dsolve.  And 2016.2 understands it as such and calculates it as it should.  Am I correct?

I wonder if anyone else could verify this 2016.1 issue.  Does it exist in 2016.0?