Search found 55 matches

by Katjes
Thu Jun 06, 2019 11:58 am
Forum: Data Manipulation
Topic: tabplace
Replies: 3
Views: 226

Re: tabplace

This was my test version - but if I would add a particular row number (not step 2) saved in an additional table (tab_row) the procedure did not work. !j=14 for !i=1 to 5 scalar row =@val(tab_row(!i)) !row =row for !i=14 to 100 step+2 %dest="A"+@str(!j) %source_a="A"+@str(!row) %s...
by Katjes
Thu Jun 06, 2019 6:23 am
Forum: Data Manipulation
Topic: tabplace
Replies: 3
Views: 226

tabplace

Is there a way to loop over a particular number of rows in tabplace, instead of providing the numbers manually. Instead of tabplace (tab_new, tab_old, "A14","A14","N14") tabplace (tab_new, tab_old, "A15","A16","N16") Something like !j=14 fo...
by Katjes
Wed May 29, 2019 3:53 pm
Forum: Programming
Topic: Forecast averaging
Replies: 2
Views: 112

Re: Forecast averaging

I now solved the issue with a loop, but was wondering whether there is a more suitable solution: 'flexible training period" genr mse_fix=NA genr mse_all=NA for !ii=0 to 31 smpl 2011q1+!ii 2011q1+!ii %train_fix="2011q1+"+@str(!ii)+"-8 2011q1+"+@str(!ii)+"-1" ''' fix...
by Katjes
Wed May 29, 2019 5:07 am
Forum: Programming
Topic: Forecast averaging
Replies: 2
Views: 112

Forecast averaging

My forecasts are recursively estimated. Therefore, I think it is reasonable to adjust the training sample for forecast averaging. Is there an easy command to do this: for %weight mse ranks smpl 2013.1 2018.1 GDP.forcavg(name=test_{%weight}, wgttype={%weight},trainsmpl="2010Q1 2012Q4") GDP_...
by Katjes
Fri Nov 23, 2018 8:12 am
Forum: Data Manipulation
Topic: Multiple stacked bar charts
Replies: 0
Views: 2867

Multiple stacked bar charts

I have a nice stacked bar chart (gr_contributions) showing contribution of a,b,c in each year for country X. Is it possible to show for each year also a stacked bar chart showing d, e,f for country Y for a particular year next to the corresponding bar of country X? group X a b c graph gr_contributio...
by Katjes
Thu Nov 23, 2017 2:23 am
Forum: Data Manipulation
Topic: Chart with different samples
Replies: 1
Views: 797

Chart with different samples

I would like to make a chart including two (or more) series, e.g. line chart. In the workfile data is available for the whole sample for all series but I would like to show different samples, e.g. x only from 2005-2010 and y from 2010-2015. Of course I can generate a new series containing only the d...
by Katjes
Thu Sep 22, 2016 2:36 am
Forum: Programming
Topic: CLS in LS equation
Replies: 0
Views: 1633

CLS in LS equation

Dear all, we used version 9.5 and version 8.1 to perform e.g. equation eq_1.ls(optmethod=legacy, arma=cls) hours c ma(1) ma(2) ma(3) ar(1) ar(2) @trend that gives us a least-square estimation (without the sigma-parameter). Both version provide the same results, Now one of my colleagues has version 9...
by Katjes
Tue Jun 14, 2016 9:21 am
Forum: Suggestions and Requests
Topic: Forecast combination
Replies: 0
Views: 1662

Forecast combination

The forecast combination command(s) works well, if all forecast series to be combined have a value, if there is an NA for forecast, no combined forecast can be calculated for the remaining series. Since I have used the group "gr_f" element to combine the forecasts it seems that I have to d...
by Katjes
Wed Jun 08, 2016 11:53 pm
Forum: Estimation
Topic: Midas equation coefficients
Replies: 1
Views: 993

Midas equation coefficients

Is there an EViews command that can be used to extract/save the elements of the lower part in the MIDAS estimations output directly?
coefficient, distribution shape,...
or is it necessary to freeze the output as table and extract the values manually?
by Katjes
Wed Nov 04, 2015 10:16 am
Forum: Data Manipulation
Topic: Histogram Overlap
Replies: 1
Views: 1143

Histogram Overlap

Hi there, I would like to plot two histograms for two series ff1 ff2 in one graph - no problem. But it didn't find a way (in EViews 9) to deal with the fact that you cannot see the 2nd histogram bins if they are smaller then the first one, and vice versa. group ff ff1 ff2 freeze(graph) ff.distplot(s...
by Katjes
Wed Jul 22, 2015 12:14 am
Forum: Programming
Topic: Forecast evalution
Replies: 3
Views: 1357

Re: Forecast evalution

Thanks for prompt reply and implementation, series.fcasteval(evalsmpl="2004 2014", testname=m_enc, statname=vv) aa bb cc in matrix m_enc the results for the forecast encompassing test (F-stat, pval) are saved. but in svector vv only the name of the first variable is given (maybe this shoul...
by Katjes
Mon Jul 20, 2015 7:48 am
Forum: Programming
Topic: Forecast evalution
Replies: 3
Views: 1357

Forecast evalution

Using EViews9, we can easily conduct forecast evaluation and encompassing tests :D series.fcasteval(evalsmpl="2004 2014") aa bb cc Is there a way to directly access the results that are shown in a spool (which is not save automatically) ? For instance to extract the values of the p-statist...
by Katjes
Mon Jul 20, 2015 7:36 am
Forum: Programming
Topic: Forecast combination
Replies: 1
Views: 593

Forecast combination

Started using the new options of EVIEWS 9, I very happy that forecast combination procedure are ready to use for indidual forecasts or a group of forecasts (ind_f1_w_all) Either manually or by command, e.g. data_real.fcastavg(wgt=mean,name=f1_av) ind_f1_w_all data_real.fcastavg(wgt=mean,name=f2_av) ...
by Katjes
Thu May 22, 2014 2:29 am
Forum: Models
Topic: add factor in models
Replies: 0
Views: 1732

add factor in models

Is it possible to set the add factor to dlog of variables as well? model m1 In a first step I calculate many equations e.g. for variable x in dlogs: smpl 1991 2012 equation eq_x eq_x.ls dlog(x) c dlog(x(-1) eq_x.makeresids res_x where the latter are the residuals of the growth rate of x Then the equ...
by Katjes
Wed Jan 22, 2014 2:14 am
Forum: Programming
Topic: Frequency conversion for sample
Replies: 1
Views: 1092

Frequency conversion for sample

I would like to convert a monthly series to a annual one, but selecting the conversion method for 1990-1995 average and for 1996 1996 the last value. I do not select the conversion method the following code works for the average part, and it works. wfcreate(page=page1) m 1990 2010 series y=nrnd page...

Go to advanced search