Comments on: Seven ways to test your number-crunching code http://www.programming4scientists.com/2008/12/22/seven-ways-to-test-your-number-crunching-code/ Software development wisdom and common-sense for the scientist-programmer Sun, 28 Feb 2010 19:34:26 +0000 http://wordpress.org/?v=2.8.4 hourly 1 By: Rich http://www.programming4scientists.com/2008/12/22/seven-ways-to-test-your-number-crunching-code/comment-page-1/#comment-109 Rich Mon, 22 Dec 2008 13:58:49 +0000 http://www.programming4scientists.com/?p=376#comment-109 I completely agree! Other people can be very creative when it comes to breaking your code and this is a Very Good Thing, because it's a form of testing :-) I completely agree! Other people can be very creative when it comes to breaking your code and this is a Very Good Thing, because it’s a form of testing :-)

]]> By: nico http://www.programming4scientists.com/2008/12/22/seven-ways-to-test-your-number-crunching-code/comment-page-1/#comment-108 nico Mon, 22 Dec 2008 13:49:53 +0000 http://www.programming4scientists.com/?p=376#comment-108 I would add: if someone else is going to use your code, have them testing it! Let them misuse the program as worst as they can... and bugs will come out! It happened to me that I wrote some routines for the analysis of a time series. Well, it worked perfectly on all my data, but when someone else in the lab tried to use it it didn't work anymore, giving weird results. Well, the problem was I was assuming (never do that!) that the time series was starting from time 0. This was always the case with my data, but not with my colleague's... I would add: if someone else is going to use your code, have them testing it! Let them misuse the program as worst as they can… and bugs will come out!

It happened to me that I wrote some routines for the analysis of a time series. Well, it worked perfectly on all my data, but when someone else in the lab tried to use it it didn’t work anymore, giving weird results. Well, the problem was I was assuming (never do that!) that the time series was starting from time 0. This was always the case with my data, but not with my colleague’s…

]]>