Search found 2598 matches

by EViews Glenn
Thu Sep 13, 2018 11:47 am
Forum: Estimation
Topic: Obtaining weights with Principal component analysis
Replies: 4
Views: 1753

Re: Obtaining weights with Principal component analysis

I must admit that I don't entirely understand what you are trying to do with your calculation. The following shows the equivalence between post-multiplying the data by the eigenvectors and the score series created using normalize loadings scaling. Note that as expected the average sum of the squared...
by EViews Glenn
Thu Sep 13, 2018 10:57 am
Forum: Estimation
Topic: Fixed Effects- Industry and Year
Replies: 17
Views: 258

Re: Fixed Effects- Industry and Year

In the language of factors, multiple terms in @expand gives the interaction effects, but not the main effects.
by EViews Glenn
Wed Sep 12, 2018 10:02 am
Forum: Suggestions and Requests
Topic: percentile function
Replies: 1
Views: 335

Re: percentile function

How about a built-in

100 * @ranks(x) / @obs(x)
by EViews Glenn
Tue Sep 11, 2018 3:54 pm
Forum: Estimation
Topic: Confidence Intervals: Forecast vs. Makemodel
Replies: 5
Views: 160

Re: Confidence Intervals: Forecast vs. Makemodel

Is there a chance you can share your workfile. In all of my test cases I'm not seeing this discrepancy. (Note that the model simulation doesn't include the coefficient uncertainty, but in my cases, I'm not seeing much difference from that).
by EViews Glenn
Tue Sep 11, 2018 1:14 pm
Forum: Suggestions and Requests
Topic: Fiscal/Academic year labeling
Replies: 2
Views: 541

Re: Fiscal/Academic year labeling

Question, you can always create an undated workfile with that kind of labeling. Is there anything that you need date wise that isn't handled by an undated with custom labels?
by EViews Glenn
Tue Sep 11, 2018 1:13 pm
Forum: Suggestions and Requests
Topic: Expected to see Eviews Workfiles Encryption
Replies: 1
Views: 442

Re: Expected to see Eviews Workfiles Encryption

It is an interesting idea and I can understand the appeal There are, however, legal issues that make this not quite as simple a request as one might think. Given this fact and the existence of third-party encryption tools which can be used to encrypt a workfile, this feature is currently not high on...
by EViews Glenn
Thu Aug 30, 2018 8:55 am
Forum: Estimation
Topic: Confidence Intervals: Forecast vs. Makemodel
Replies: 5
Views: 160

Re: Confidence Intervals: Forecast vs. Makemodel

We are looking into this for you. It may take a bit but we will let you know what we discover.
by EViews Glenn
Tue Jun 05, 2018 10:18 am
Forum: Data Manipulation
Topic: factor based on deviation from mean
Replies: 1
Views: 313

Re: factor based on deviation from mean

You want the comparison to the mean of the other variables for each observation?
by EViews Glenn
Mon Jun 04, 2018 6:27 pm
Forum: Data Manipulation
Topic: Copy values from cross-section to entire vector
Replies: 3
Views: 279

Re: Copy values from cross-section to entire vector

Code: Select all

series var2a = @meansby(var2, country, date)
by EViews Glenn
Fri Jun 01, 2018 4:24 pm
Forum: Data Manipulation
Topic: Copy values from cross-section to entire vector
Replies: 3
Views: 279

Re: Copy values from cross-section to entire vector

Can you be a bit more specific about what you are doing? What values in AA need to be copied? Is it the value for one date, or values for a range of dates? Are we matching up observations by-date?
by EViews Glenn
Thu May 31, 2018 11:13 am
Forum: Estimation
Topic: Second Order Differencing
Replies: 7
Views: 1248

Re: Second Order Differencing

Forecast of data at t+4 given information at t.
by EViews Glenn
Thu May 31, 2018 11:12 am
Forum: Estimation
Topic: State Space - Prediction Intervals?
Replies: 1
Views: 205

Re: State Space - Prediction Intervals?

You can request the forecast standard errors and construct intervals from those.
by EViews Glenn
Thu May 31, 2018 11:09 am
Forum: Data Manipulation
Topic: outliers in a scatterplot
Replies: 1
Views: 193

Re: outliers in a scatterplot

If you are okay with a separate regression line for both the non-2008/2009 values and for the the 2008/2009 values, the easiest way to do this is to create new series that have missing data for subsections of the data. That is, create a series with just the 2008 and 2009 values, and a separate serie...
by EViews Glenn
Fri May 25, 2018 4:47 pm
Forum: Bug Reports
Topic: Cointegration Test gives wrong P-Value
Replies: 3
Views: 298

Re: Cointegration Test gives wrong P-Value

We've identified the issue. In the MacKinnon algorithm for the response surface, we have found that there are potential issues with the polynomial approximation to the response surface for extreme values. In this case, the statistic value is so out of bounds that the existing algorithm approximation...

Go to advanced search