Print

Print


Dear Jurgen and Jaap,

Thanks kindly for the helpful suggestions. I am not sure what my problem is. In excel, my dates column is the first column and it does not have a title; while the my observations column has a title. Also, I entered the dates as monthly as, eg., 1980-1 (although when I hit enter, this was converted to Jan-80 instead of remaining as 1980-1). Jurgen is correct, I had mistakenly typed 'deterministic' instead of 'Deterministic'. But after correcting this mistake, only the problem of 'not a function member' was corrected, the time series type problem did not go away.

I shall revisit the dates issue in excel, as per your suggestions.

Sorry I had been away from my computer, hence the delayed response.

PS: Jurgen I do not have the OxMetrics book. I only have the 'Ox Introduction' pdf manual. Do I have to buy the OxMetrics book from Timberlake Consultants? If so, I shall plan to do that. In fact, I am budgeting to buy Professional Ox because I think I am beginning to understand and like Ox.

Best regards,
Lexi

On Thu, Jun 30, 2016 at 9:19 AM, Jurgen Doornik <[log in to unmask]> wrote:
Dear Lexi,

As I wrote yesterday, the problem is probably with the way that the dates are recorded in the xls file. The easiest way to see a file that works is to look at a file that works, e.g. ox/data/data.xls.

From section 11.2 of the OxMetrics book:

The data matrix is a rectangular array, structured in one of the following formats:
1. Unlabelled first column with dates, followed by labelled columns, one for each variable.
The first column either consists of proper calendar dates, or strings that represent a date. In the latter case, it should be of the form year–period, where the – can be any single character: for example, 1980–1 (or: 1980Q1 1980P1 1980:1 etc.; this character is ignored).
If the date column can be read as fixed frequency dating of observations, the
sample will be set appropriately, and the date column will not appear as a variable.
If the first column cannot be interpreted as defining a fixed-frequency database, the column will be loaded with default variable name Svar1.

2. All columns are labelled.
An attempt will be made to read the first column as the dates. Regardless of the
result, the column will appear as a variable in the database.
If string-based dates were used, the date variable is of type choice. Otherwise it will be of type date, in which case the database will appear as dated.

3. No column is labelled.
This is the same as if all columns are labelled with Svar1, Svar2, ....
If no dates are present, the database will have a fixed frequency of unity, starting from ‘year’ 1.

Best wishes, Jurgen

oxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxox
oxoxoxox 18th OxMetrics user conference 2016
oxoxoxox      Cass Business School, 12-13 Sept
oxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxox
Dr Jurgen A Doornik
James Martin Fellow, Institute for New Economic Thinking
at the Oxford Martin School, University of Oxford
http://www.doornik.com
http://www.oxmetrics.net
oxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxoxox

On 2016-06-29 23:26, Vries, J.J. de wrote:
May be *Resample* helps to set your database as monthly data.

<http://www.doornik.com/ox/index.html?content=http://www.doornik.com/ox/oxclass.html#Sample::Resample>

greetings,
Jaap
--------------------------------------------------------------------------------
*Van:* The ox-users list is aimed at all Ox users [[log in to unmask]]
namens Lexi Setlhare [[log in to unmask]]
*Verzonden:* dinsdag 28 juni 2016 22:18
*Aan:* [log in to unmask]
*Onderwerp:* Database Type

Dear Ox Users,

May I please ask for your help concerning "database type", assuming that I am
using the correct terminology.

The issue is that I have my dataset in Excel. Using my OxEdit in OxMetrics7, I
open the Excel data in ox console, using database library. Then I use the
'GetVar' command to access the variable which I model using an autoregressive
model (the purpose is to estimate persistence). After getting the variable,
called Tcore, I call rename it to be TrimP and then I create its lagged term and
call it TrimP1.

I then append TrimP and TrimP1 to the original dataset; after which I select
TrimP as dependent variable and TrimP1 lagged three times as regressors.

The problems are as follows:
(i) Although my data is in excel and at monthly frequency, my ox program does
not recognize it as time series. I say this because when I try to create
deterministic terms, I get an error saying 'deterministic not a function
member'. Also, the .Info() function gives
---- Database information ----
Sample:    1 - 116 (116 observations)
Frequency: 1
Variables: 4

Please assist me to get ox to recognize my data as time series. My ox program is
attached.

Thanks,
Lexi