4 ms·
Wikipedia is magnificent for actually learning the algorithms, but I think collections like these fill a gap in that they show implementations in actual program
by johnsonjo 8y ago
Wikipedia is magnificent for actually learning the algorithms, but I think collections like these fill a gap in that they show implementations in actual programming languages that just work without trying to write code from pseudo code. If you’re familiar with any of these languages and understand the theory backing the algorithm you can instantly see how these algorithms can be done in that target language.
I feel this is especially useful for me with the computational geometry part. I took a completely theoretical computational geometry class this last semester. We learned and made algorithms in pure pseudo code, and some operations we were never even given pseudo code for, because they were ‘trivial’, for example the algorithm for line intersections (my teacher at least told me it took some linear algebra, but I never took linear algebra.) Having just the code can be helpful for someone that already has most the theoretical backing, but needs the practical code to go along with it. I think this kind of practical code along with getting theory from a more theoretical book or Wikipedia entry is a powerful way you could learn the algorithms.
- gepoch 8y agoHave you looked at Rosetta code before? I think OP has a pretty good point bringing it up.. http://rosettacode.org/ http://rosettacode.org/
- johnsonjo 8y agoI've come across it before, but haven't really used it. It looks like the examples are pretty good. I however dislike how it is organized as it organizes by first letter of a title (that seemingly has no pattern for first words making finding a particular entry harder) and doesn't have different categories or tags to categorize the different types of tasks. Organizing by type of task would have been a nice feature for browsing the code I much rather enjoy the layout of the linked github as it's much easier to study a particular topic.