liberty-eiffel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Liberty-eiffel] SoC 2016


From: Bernd Schoeller
Subject: Re: [Liberty-eiffel] SoC 2016
Date: Sun, 6 Mar 2016 19:44:06 +0000
User-agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.10; rv:38.0) Gecko/20100101 Thunderbird/38.6.0

Just to add to this, I am currently implementing a MATRIX class and related linalg function for the other Eiffel compiler. I am trying to build something like numpy, but - of course - with a better design and consistent and powerful interface.

https://github.com/schoelle/algae

I am not saying that you should use this: it is probably not much further developed than starting from scratch and the two Eiffel compilers do differ quite a bit.

But perhaps we can have an open exchange on design decisions, as my stuff is also still strong 'work in progress' and I have no clue how I want to implement certain things like the different matrix decomposition functions and linear regression.

Good luck,
Bernd

PS: A book that I can very much recommend:

https://books.google.co.uk/books/about/Matrix_Computations.html?id=5U-l8U3P-VUC

I found the 3rd edition here for free:

http://web.mit.edu/ehliu/Public/sclark/Golub%20G.H.,%20Van%20Loan%20C.F.-%20Matrix%20Computations.pdf

On 06/03/16 18:49, Raphael Mack wrote:
Hi Alyona,

this is great. We are looking forward to your concrete proposal.

For the Scientific Wrappers you should be able to show substantial OO
knowledge, maybe you have been involved in some library design and can
give references. You should search the web for "what a programmer should
know about floating point" and the first 3 results should not confound
you and the lectures "linear algebra" or "computational science" shall
not be the worst in your memorization. - A very interesting topic by the
way - in a world, where math is not fully respected (due to inaccuracy)
we are still able to produce meaningful results.

Windows support is probably easier, and less OO and Eiffel knowledge is
necessary but without C experience it will also get hard. At least one
toolkit to create a windows "setup" should directly come to your mind
and most of the topics "low level C programming", "network programming",
"controlling external processes", "scripting", "cygwin", "ming-w" should
mean something for you.

But don't be afraid, you will not be alone - your mentor is there to
help, and if you are motivated it will be fun, you will have success and
make good experiences.

It would also be interesting so see how you think that your work can
help LibertyEiffel even beyond the concrete work product you have at the
end of the summer. (Applies to all GSoC aspirants.)

To sum up: Beside a non-negligible amount of time during the summer and
some passed exams in a Computer Science (or similar) program the mean
prerequisite is motivation. Please show this in your proposal.

Cheers,
Rapha

Am Sonntag, den 06.03.2016, 13:33 +0100 schrieb Alyona Morozov:
Hi everyone,


I am Alyona Morozov, pursuing a BSc in Computer Science from the
University of York, England. I am a 2016 GSoC aspirant and would like
to do
a project on LibertyEiffel, developing the Scientific Wrappers or
Windows support project for Liberty Eiffel. I am currently looking at
the code on Github. I am planning on starting work on these proposals
this week and submit to a mentor for early review.


I will like to know if there are any prerequisites to contribute to
this project?

Thank you for your reply,

Sincerely,
Alyona







reply via email to

[Prev in Thread] Current Thread [Next in Thread]