I’m going to buy a package for econometric modelling. From what I have read - seems EViews is popular. But what about the others? - GAUSS - MATLAB - RATS - SAS - SHAZAM - SPLUS - SPSS - TPS I’m mot a techie - won’t do coding from stratch - so i need windows, pre-built calcs for most things. Can do most things in Excel (building own formulas, or using some built-in functions), but I need it to do things better/quicker for serious crunching - eg Regression stats, Jarque-Bera, Fisher exact, Levene, White, Durbin-Watson, Breusch-Godfrey, Ramsey, Chow, Box-Jenkins, Goldfield-Quandt, Hannan-Quin, Box-Pierce, Ljung-Box, ARMA, VAR, unit roots, Engle-Granger, Dickey-Fuller, Johansen, ARCH, GARCH, EGARCH, VECH, BEKK, Markov, logit/probit, Monte Carlo - that sort of stuff Is EViews best for this type of stuff - or is there a better package? cheers…
Gretl is pretty good for a free jobbie - does most of your list, and provides links to R for all the stuff it can’t do. But that means knowing how to program in R http://gretl.sourceforge.net/
hi chris - thanks for the tip. I will also look into PSPP - which is a free imitation of SPSS, I think. Maybe a free one may be the way to go initially, even if not as comprehensive as the expensive packages… btw - did you make it down here over???
It seems you are going to need it mainly for statistical purposes. Eviews will suffice in most of the cases and SPSS is pretty convenient for point and click but if you need speed, efficiency and capability of handling enormous amount of data, then SAS is your best bet.
I think Eviews is very aestheticallly pleasing but I’m not sure it will always get the job done if you are working with complex models
Came over for Christmas/New Year round brissy - had a very expensive but very enjoyable New Year at Hamilton Island Still contemplating coming out full time - is your number still ******8003?
yup - you should have called - I have a couple of houses up there…!
2x2equals4 Wrote: ------------------------------------------------------- > It seems you are going to need it mainly for > statistical purposes. Eviews will suffice in most > of the cases and SPSS is pretty convenient for > point and click but if you need speed, efficiency > and capability of handling enormous amount of > data, then SAS is your best bet. 2x2 - I’ve read that SAS is tremendous but that it’s more or less like a coding langauge. But what if i am a real dumb-a$$ - is it easy to use with pre-built formulas, like EView looks like from the demos? thanks…
Bummer! My pommie reserve clearly not working well for me! Stunning place. Managed to burn out our golf cart though
EViews is definitely the easiest engine to use, but is also the most basic. You should probably spend the time learning SAS or STATA because those are the programs that are used in corporate environments.
NN wants nothing to do with the corporate environment…trust me
nothing wrong with corporates - I was there once. Now I advise corporates, fund managers, fuondations, etc - they are still trying to keep themselves in yachts & bentleys. Poor buggers - riddled with debts, spending more than they earn, wives/kids/cars/boats/houses/etc they can’t afford. Recessions are great from time to time - a good clean-out of all the crap so we get back to basics. Yachts & Bentleys are soooooooo Grenspan era! It’s a new world now… cheers…
MatLab has most of the in-built functions that are in your list. Its not hard to code, could be learnt within a couple days time. R/SPLUS is very easy to code, and I mainly use it for statistical analysis(all the test, ranks tests, paired sample tests n all). You can do some finance stuff on R like time series n all but I dont know many people who use R for finance. SAS, C++ , Java and SQL these are the ones that are commonly used industry wide, I use them as well. Apart from basic regression analysis in school I dont remember using Eviews for anything else.
Stata has pretty much everything you need. What’s not built-in can be added with a do file.
Stata is quite expensive, isn’t it?
TheAliMan Wrote: ------------------------------------------------------- > Stata is quite expensive, isn’t it? Yes - unless you get it through a university deal, v11 with full documentation is well over a grand. Even with the university discount, it runs over $800.
All the packages will be fine, if your only issue is price then get the matlab ® equivalent which is free, or if you have some spare cashola fork it out. I myself use R but will usually compare stats against what Eviews computes… the great and worst thing with the latter is you’re not witness to what it does once you mouse click BTW many of those applications are seeded on torrent sites.
I wanted to suggest that, but didn’t know if it would be bad for business purposes. Is R the free version of matlab, or are you referring to something else? I spent like ten minutes looking at the manual awhile back and it seemed like a pain in the ass compared to matlab.
nope i was referring to Octave but i don’t think the OP is really looking at a number cruncher
I like using Matlab because it’s very powerful and flexible. You can install free econometrics libraries (that are better than the official econometrics toolbox) from http://www.spatial-econometrics.com and http://www.kevinsheppard.com/images/2/2d/Ucsd_garch.zip.