The RATS Software Forum

If you have suggestions for features or improvements you would like to see in the forthcoming Version 7 of RATS, please send them to us at:

support@estima.com

Thanks!

You are looking for suggestions for the coming version 7.

Here are a few suggestions to expand the graphics capabilities of RATS:
o Split graphs (without any space)
o Draw horizontal lines
o Choice of color
o Control of line thickness
o Support for more graphics format
o Footer in basic chart (GRPARM is working only with SPGRAPH, charts need more space)
o Shading in different colors
o ...



Then I have another suggestion concerning data reading. I am working with very large databases with DAILY periodicities. If I want to read a series, I have to specify the name of the RATS data file! Would it be not possible to open a large number of data files simultaneously and RATS would figure out the correct data file. At the moment, I have to create a huge RATS datafile to put all the different RATS data files together(Something similar to FORMAT=HAVER or FORMATO=DRI).


Do you have a planned release date for version 7?

Regards
Alex

Alex:

Thanks for the suggestions. Some of these are already in the works. I did have a couple of questions.

First, note that you can already choose the color used for lines/patterns from a set of predefined choices by using the "symbol choice" parameter on the supplementary card(s). Is this what you wanted, or were you thinking of having the ability to completely select your own colors (by specifying RGB values, for example)?

Second, what exactly do you mean by "split graphs"? Could you describe this in more detail?

Finally, are there any particular graphics formats you would like us to support?

As for the data file suggestion, we can give that some thought. However, if you have a collection of large files with a lot of series, it may be a bit slow to have RATS scan through all the files to find the series you want.

No set date for Version 7 yet. We'll be releasing a final minor version 6 update shortly, with 7 to follow in a few months.

Thanks again,
Tom Maycock
Estima

Hi Tom

Many thanks for your answer.
- "split graphs", I will send you an example today via email
- graphics format: *.gif and *.jpeg formats (*.wmf format is using a lot of memory)

I have another suggestion: With a two-scale graph, it would be useful to plot one time-series "in reverse mode" (the highest at the bottom and the lowest at the top).

Concerning my data file suggestion, I think about another solution with the command "CATALOG" (reading of all the files and writing the var information into a file).

Regards
Alex

HinderAlex wrote:Hi Tom

Many thanks for your answer.
- "split graphs", I will send you an example today via email

Thanks, as noted in my email:

You can already do this to some extent using the "OVRANGE" option (in conjunction with OVERLAY and related options) on GRAPH. You specify a fraction indicating the degree to which the two sets of series (one for the left scale and one for the right scale) overlap vertically.

For example:

cal 2000 1 12
all 2007:5
set y = 10 + %ran(1)
set x = 20 + %ran(1)
graph(overlay=line,ovrange=.4) 2
# y
# x

gives a graph where the two series are separated vertically by a fair bit of white space (the series associated with the left scale is graphed in the bottom 3rd or so of the graph space while the series associated with the right side scale is graphed in the upper 3rd of the graph space).

Yes, please include:
-- Markov switching SVAR (SZ 2006 and RWZ 2006)
-- Markov switching BVAR (SWZ)
(Matlab codes for these two are apparently available from Tao Zha's webpage);
-- also, MS-VECM

Thanks! Also, impulse responses for all these and MS-VAR and graph of probabilities as in SZ and RWZ papers. Lastly, is there an easy way to do Kim (J Econometrics 1994) and Kim-Nelson (1999) dlm with m-s stuff? (I'm novice RATS user). I know all these stuff are worth more than what I paid for, but thanks if these can all be done!

moderator wrote:If you have suggestions for features or improvements you would like to see in the forthcoming Version 7 of RATS, please send them to us at:

support@estima.com

Thanks!