yes, i have triumphed over my nerve-wracking biometry assignment. well, i've triumphed over problem numero uno, anyway. and that's enough triumph for me for one day. i've spent over 4 hours of my day (i started a little after 8, and i just finished around 12:30) teaching myself how to write a program in minitab to do something that i didn't really understand in the first place. no, the macros in our notes did not help me out much. but i used them anyway. which is probably why the stupid thing refused to run the first eight hundred times i tried it...for 800 different reasons. stupid minitab. stupid mccreadie. stupid randomization. one program down, two more to write. but not today, baby...not today. i'm glad i didn't go to class today. if i had, i probably wouldn't have written a working program this morning. anyway, i wrote the program. it works. the results are comparable to the results i get with traditional statistics (the easy way - when i just type in "corr c1 c2" and minitab spits out a nice correlation coefficient and a p-value), which according to mccreadie means it's right. if the results are the same, just what the hell is the point of doing 1,000 randomizations (i'm not exaggerating this time. the program literally has to do 1,000 randomizations of the data)? anyway, it's done, and i did it all by myself! i'm like a little kid who's just learned to use the potty. i'm so proud of myself. and now i'm one problem closer to being finished for the semester...
1 comment:
Congratulations on getting your program to work!
I know how you feel ... I feel like I've won the Stanley Cup, or exorcised a demon, whenever I write a program that works. When I was one of the finalists for my job at Gulf Power, I was asked to fix macros for Excel. It took me six hours to do it, and I had no idea what I was doing, but I did it.
And the rest is history.
I may want to see your work some time.
Post a Comment