GSoc 2014:Proposal of a new idea

[Rishabh Mishra]

Respected Sir,

I am Rishabh Mishra,2nd yr undergrad B.E.(hons.) Chemical Engineering student from Birla Institue of Technology and Science Pilani,Goa Campus.I am new to python,but have a very sound background of java and c.I recently got introduced to see about 4 months ago and was fascinated by how easy it makes the code to write.

ABOUT THE PROPOSAL:As i said i am new to python so i don't know what all packages are there in sympy but by my research there is nothing there on fluid dynamics or Mass transfers (for chemical industries).So. i would like to undertake a project on fluid dynamics (which i have already completed with a grade of 8/10) or mass transfer (which i will finish by this summer).

Please reply what you have in mind, and if you are interested i will put up with what all ideas i have in mind for the particular field.

[Manoj Kumar]

Hi Rishabh,

I had a really similar idea, and I would love to help you (though I cannot mentor this summer)

1. The very little amount of heat/mass transfer that I've just read to pass my grades requires solving for PDE's or ODE's numerically, if I'm not wrong. Could you please be more specific as in what exactly you would like to implement? Fluid Mechanics is a really wide area, could you again be more specific?

And the other question is (@Aaron, Ondrej and other core developers) does this thing fit into SymPy? Would we like to have a sympy.thermodynamics or a sympy.fluid library?

Thanks.

[Tim Lahey]

Solution of ODEs and PDEs numerically fits with SciPy, not Sympy.

sympy.physics.fluids and sympy.physics.thermodynamics would only make
sense for modules that support the derivation of equations in those
fields. That's what the mechanics and the quantum physics modules do.

Cheers,

Tim.

---
Tim Lahey, Ph.D.
Post-Doctoral Fellow
Systems Design Engineering
University of Waterloo

[Rishabh Mishra]

ThankYou for the reply.

I want to add libraries for directly solve eqations of navier's stoke and other's using some standard data, ad calculation of pressure drops for a flow across the pipe.

As mentioned by Tim, if this doesn't fall under SymPy then maybe i should move this discussion to SciPy.

[Rishabh Mishra]

Can anyone please help me where should i discuss my proposal.Thank you.

On Friday, March 14, 2014 11:47:27 PM UTC+5:30, Rishabh Mishra wrote:

[Tim Lahey]

Solving Navier-Stokes involves solving a system of PDEs numerically, so
that falls outside Sympy. As for calculation of pressure drops for flow
across a pipe, those aren't even close to enough for a GSoC project.

To be honest, Navier-Stokes doesn't really fit too well with SciPy as
it's kind of specialized. It does fit with SfePy (http://sfepy.org/). I
don't know if they're doing GSoC, though.

Cheers,

Tim.

> --
> You received this message because you are subscribed to the Google
> Groups "sympy" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to sympy+un...@googlegroups.com.
> To post to this group, send email to sy...@googlegroups.com.
> Visit this group at http://groups.google.com/group/sympy.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/sympy/df380ca5-c7e1-49e1-8ec9-c1d31a4f1d7d%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Jason Moore]

Rishabh,

Your best bet is to look through the ideas pages of different organizations and choose an idea that has already been suggested.

Jason
moorepants.info
+01 530-601-9791

--
You received this message because you are subscribed to the Google Groups "sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sympy+un...@googlegroups.com.
To post to this group, send email to sy...@googlegroups.com.
Visit this group at http://groups.google.com/group/sympy.

To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/af50db09-79c6-4b21-90d3-d1ed3ab4d369%40googlegroups.com.

[Rishabh Mishra]

As i have found that there may be no application for navier stoke and pressure equation.

I have come up with a new idea of solving polynomial equations using various methods such as newton raphson,brent's,bairstow's and muller's method.

Can anyone please help me with how feasible the idea is,or is the out of the sympy domain as well (as far as i have seen there are functions for solving polynomial equations,but still) .

Thank you.

On Friday, March 14, 2014 11:47:27 PM UTC+5:30, Rishabh Mishra wrote:

[Jason Moore]

Rishbabh,

You are likely too late in the game to try to formulate a new idea for SymPy. If you want to apply to SymPy please review the ideas page (https://github.com/sympy/sympy/wiki/GSoC-2014-Ideas) and find one of the many ideas that interests you. Then learn as much as you can about it from past email discussions, past pull requests/issues, and asking very specific new questions to the email list. Good luck.

Jason
moorepants.info
+01 530-601-9791

--
You received this message because you are subscribed to the Google Groups "sympy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sympy+un...@googlegroups.com.
To post to this group, send email to sy...@googlegroups.com.
Visit this group at http://groups.google.com/group/sympy.

To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/b999244e-f393-4d1a-ae72-7adbac879ab7%40googlegroups.com.

[Joachim Durchholz]

Am 15.03.2014 18:20, schrieb Rishabh Mishra:
> As i have found that there may be no application for navier stoke and
> pressure equation.
> I have come up with a new idea of solving polynomial equations using
> various methods such as newton raphson,brent's,bairstow's and muller's
> method. Can anyone please help me with how feasible the idea is,or is the out of
> the sympy domain as well

Those are all numeric (i.e. approximative) methods.
SymPy is not for that area, it's for symbolic (i.e. precise,
closed-form) math.

> (as far as i have seen there are functions for
> solving polynomial equations,but still) .

These are for finding precise roots I suspect.
Any numeric code that you could find in SymPy would be there to support
something symbolic, and have special justification.