Finally, I have started using R. I sort of waited until I had something to start from scratch, and changing computers helped cos now I’m using SAS on the server and it sucks.
So far I’ve opened a database – nothing else can be done otherwise (thank you XL for putting commas as separator in the csv file when you specify you've used semi-colomn)! Made scatterplots, correlations, descriptive analyses, and now have moved to ordinal logistic regression. Have found how to check for collinearity. Need to understand the assumption checking for ordinal logistic still.
I’d like to find out if it’s possible to edit large series of commands and run them at once (as would be possible in SAS for example) instead of just running a few at at time, but otherwise I’m finding my way around.
I’ve done all the data editing in a csv file in XL, which is only allowed because I don’t have a gigantic database, but it’s quite handy.
Because I’m used to that, and because I’m fairly sure I won’t have everything in those drop down menu, I’m not using Rcommander.
This was a great help to get started, and this one remains a classic, whatever software you use. Good for R too.
Any one else has made the jump to R?
Thursday, 2 July 2009
Subscribe to:
Post Comments (Atom)
3 comments:
I've thought about it because I know that before too long I may not be in a place to afford SAS.
Thanks for including the links.
darnit, you beat me to it! I might just have to catch up with you at some point!
Good luck to you both in starting if/when you do. It's really not impossible if you're familiar with something like sas; but I suspect that start at the beginning of a piece of a work helps. Also you might want to check out R commander, with nice drop-down, clickable menus. Didn't use it much myself though. My brain is set on 'keyboard commands' by default when it comes to statistics...
Post a Comment