Difference between revisions of "Code"
(→Tutorials) |
|||
| Line 1: | Line 1: | ||
This page contains a list of codes we've found useful for scientific HPC from a developer's perspective. To use existing codes, check out [[Best Practices]]. |
This page contains a list of codes we've found useful for scientific HPC from a developer's perspective. To use existing codes, check out [[Best Practices]]. |
||
| + | |||
| + | == Resources == |
||
| + | |||
| + | These sites allow you to run code interactively in your web browser. They are useful for trying out new ideas and following along with tutorials. |
||
| + | |||
| + | * [https://repl.it/languages/python3 Repl.it] (this is the only one that currently supports numpy) |
||
| + | * [http://www.tutorialspoint.com/codingground.htm CodingGround] |
||
| + | * [http://codepad.org/ CodePad] |
||
| + | * [http://pythonfiddle.com/ PythonFiddle] |
||
== Tutorials == |
== Tutorials == |
||
Revision as of 13:46, 15 December 2016
This page contains a list of codes we've found useful for scientific HPC from a developer's perspective. To use existing codes, check out Best Practices.
Contents
Resources
These sites allow you to run code interactively in your web browser. They are useful for trying out new ideas and following along with tutorials.
- Repl.it (this is the only one that currently supports numpy)
- CodingGround
- CodePad
- PythonFiddle
Tutorials
When learning a new language, once you've memorized the words, you need to start conversing. Unfortunately, source code is not spoken. It's a stillborn language. One of the best ways to code well is to study and modify other, good code. For code, this means reading and tweaking tutorials first, then design patterns, then large-scale projects next.
- Exercism Instructions for New Developers
- Bash By Example
- Data Structures in Python
- Linear Algebra in Numpy
- Liquid Structure With Gromacs
Remember that it's not enough to just read them. It's actually a positive waste of time. Rather, you have to do some run-tweak-run-debug development cycles with them.
- Code:realpole Real-space quadrature implementation of multipole electrostatics.
C Libraries
With the exception of the standard libraries and the GNU Scientific Library, there are surprisingly few of these about. Here are some generics that we've implemented and are sharing under the GPL.
Robust map and red/black tree code has recently been published to github.
The amazing, smallish, extensible, final-form, lambda interpreter in C [1].
Python Libraries
- libnwchem A patch against Nwchem 6.5 to build its functionality as a shared library, along with glue code for access using python ctypes.
- Networkx a quick guide.
Geometric Algebra
The Multivector class implements geometric algebra using Python's operator overloading. It works with minimal space for any number of spatial dimensions, since it's not based on the matrix / spinor representation of Clifford algebra. Instead, it takes the basis blade approach recommended by Dorst, Fondijne, and Mann [2]. It's really just as easy as ordering all the basis vectors, and keeping track of the commutator (+/- <math>\hat 1</math>).
Templates
The real world has forms, computation has templates. Pick one up, fill in the blanks, and send it down the chute.
Efficiencies
To chop a tree quickly, spend twice the time sharpening your axe.
Code Humor
Put this into your makefile: <source lang=bash> Clint\ Eastwood: myday echo "I feel great!" </source>
