I'd appreciate your advice

Jump to navigation Jump to search

I'd appreciate your advice

Edited by author.
Last edit: 17:11, 22 August 2012

I'm interested in doing a quantile regression of my products based on 4 independent variables. It's been years since stats classes in college and I've never used R. Do I have a statistically significant chance at succeeding?

Occasionaluse (talk)15:31, 22 August 2012

You should see what you can teach "Conservative" at Conservapedia about stat. His latest response to me about him bullying someone over regression analysis (of all things) shows just what an ignoramus he is and how deeply in "intellectual bunny hole" he is.

Nate Keaton (talk)16:16, 22 August 2012
 

We are talking conditional expectations: the probability to succeed is quite high provided you have some knowledge of statistics and are willing to do a little programming.

There is a package named quantreg which seems to include all the functions you need. And here is a description how it works (though I think the example is for a single independent variable. However the model can easily be extended to y ~ x1 + x2 + x3 + x4...)

If you have more detailed questions, feel free to ask!

larron[[User_talk:LArr18:01, 22 August 2012

Yeah, I was a little confused by their choice of example. I'm sure I will have a question or two. Thanks a lot.

Occasionaluse (talk)18:57, 22 August 2012
 

So I'm going through R and trying to brush up on my stats. R's not too bad. It's amazing how it gets really powerful really fast. It's the stats I'm struggling with. Could message you with some specifics of my goal and you could tell me if I'm going in the right general direction?

Occasionaluse (talk)22:20, 24 August 2012

As I said: feel free to ask - you can use e-mail this user.

larron (talk)06:35, 25 August 2012