Receiver operator characteritics

For technical questions regarding estimation of single equations, systems, VARs, Factor analysis and State Space Models in EViews. General econometric questions and advice should go in the Econometric Discussions forum.

Moderators: EViews Gareth, EViews Moderator

Alan.R.
Posts: 10
Joined: Fri Apr 29, 2016 1:27 am

Receiver operator characteritics

Postby Alan.R. » Wed May 25, 2016 2:39 am

Dear forum users, I have an equation where the dependable variable is a dummy so I perform a logit regression. I have to main independable variables and some controlable variables. I need to estimate what % of the time my model predicts correctly and also see the receiver operator characteristics: http://gim.unmc.edu/dxtests/roc3.htm
I am using EViews 9. Please help.

Thank you in advance

EViews Glenn
EViews Developer
Posts: 2682
Joined: Wed Oct 15, 2008 9:17 am

Re: Receiver operator characteritics

Postby EViews Glenn » Wed May 25, 2016 10:46 am

There's a predict view off your binary equation (View/Expectation-Prediction Evaluation) which computes various summary measures.

Unfortunately, ROC isn't built-in, but there's a forum discussion which will allow you to produce the graph with only a bit of work.

http://forums.eviews.com/viewtopic.php? ... &hilit=roc

Alan.R.
Posts: 10
Joined: Fri Apr 29, 2016 1:27 am

Re: Receiver operator characteritics

Postby Alan.R. » Thu May 26, 2016 12:58 am

Sadly I have no idea with creating EViews programs or coding...is it okay if I tell you the name of the variables and you create the code yourself?

EViews Gareth
Fe ddaethom, fe welon, fe amcangyfrifon
Posts: 13603
Joined: Tue Sep 16, 2008 5:38 pm

Re: Receiver operator characteritics

Postby EViews Gareth » Thu May 26, 2016 1:15 am

That is outside the scope of what we can provide here.

Alan.R.
Posts: 10
Joined: Fri Apr 29, 2016 1:27 am

Re: Receiver operator characteritics

Postby Alan.R. » Sun Jun 05, 2016 3:46 am

That is outside the scope of what we can provide here.
Is there a way to get EViews to produce a graph of the prediction power of my model. According to the results, with no model, if we pick at random the chance that we guess correctly is 32,15 %. WIth a model I am correct 52,58 %. Can the program produce a graph?


Return to “Estimation”

Who is online

Users browsing this forum: No registered users and 2 guests