Ok, thanks Chris!
One more question, in order to visualize the quality of my forecasting I am calculating scalars involving the forecasted and actual series such as MAPE(%) as @mape(actual,forecasted).
Any other one appart from @rmse that you would recommend?
Thanks again for the help.
Search found 14 matches
- Sat Aug 05, 2017 4:01 pm
- Forum: Estimation
- Topic: ETS Forecasting high Root Mean Squared Error
- Replies: 4
- Views: 4376
- Sat Aug 05, 2017 3:46 pm
- Forum: Estimation
- Topic: ETS Forecasting high Root Mean Squared Error
- Replies: 4
- Views: 4376
Re: ETS Forecasting high Root Mean Squared Error
Hi Chris,
The mean of my series is 28231. Is not supposed RMSE to fall between 0 and 1? Or at least that's what I found among other examples of ETS.
Thanks,
The mean of my series is 28231. Is not supposed RMSE to fall between 0 and 1? Or at least that's what I found among other examples of ETS.
Thanks,
- Sat Aug 05, 2017 3:21 pm
- Forum: Estimation
- Topic: ETS Forecasting high Root Mean Squared Error
- Replies: 4
- Views: 4376
ETS Forecasting high Root Mean Squared Error
Hi Eviews team, I performed a ETS month-ahead daily demand forecast using daily values from May 2016 in order to predict daily values of June 2016. The output looks very good when I plot the series on top of the actual values. Also the MAPE I calculated manually shows 2.59%, which is very good. But ...
- Wed Jun 28, 2017 7:53 am
- Forum: Data Manipulation
- Topic: Date Arithmetic
- Replies: 5
- Views: 8270
Re: Date Arithmetic
In case someone might find it useful, although for sure there are easier ways to program this. I just wanted to leave here what I have done in order to: - Create dummy variables for 6 UK Bank Holidays (does not include Good Friday and Easter Monday as they don't follow any arithmetic as far as I am ...
- Wed Jun 28, 2017 7:37 am
- Forum: Data Manipulation
- Topic: Date Arithmetic
- Replies: 5
- Views: 8270
Re: Date Arithmetic
Hi,
Finally I used a more complex solution I found in another post:
series bholidays = (@month=5 and @weekday=1 and @datediff(@dateadd(@datefloor(@date, "m"),1,"m"), @date, "d")<=7)
But @holidays is definetly better indeed.
Thanks,
Finally I used a more complex solution I found in another post:
series bholidays = (@month=5 and @weekday=1 and @datediff(@dateadd(@datefloor(@date, "m"),1,"m"), @date, "d")<=7)
But @holidays is definetly better indeed.
Thanks,
- Wed Jun 28, 2017 7:07 am
- Forum: Data Manipulation
- Topic: Date Arithmetic
- Replies: 5
- Views: 8270
Re: Date Arithmetic
Hi, Im using the following date command in order to create a dummy variable for Spring Bank Holiday (last Monday of May): series bholidays = (@month=4 and @weekday=1 and @datediff(@date,@datefloor(@date,"m"),"w")+1=4) But it only works when May has 4 Mondays. Therefore, year 2006...
- Tue Jun 27, 2017 12:31 pm
- Forum: Data Manipulation
- Topic: Date Arithmetic (irregular workfiles)
- Replies: 4
- Views: 6564
Re: Date Arithmetic (irregular workfiles)
Spot on Gareth!
Thanks a lot.
Thanks a lot.
- Tue Jun 27, 2017 11:41 am
- Forum: Data Manipulation
- Topic: Date Arithmetic (irregular workfiles)
- Replies: 4
- Views: 6564
Re: Date Arithmetic (irregular workfiles)
Thanks a lot for this material Chris! Following into this topic: Say if I want to create a dummy variable =1 when a date is bank holidays (see 26th of December). I would type: series bholydays= (@month=12 and @day=26) But ... how would I tell Eviews to make it =0 when that day of the week is either ...
- Wed Jun 21, 2017 1:41 pm
- Forum: Data Manipulation
- Topic: Creating New Series using conditional
- Replies: 2
- Views: 3215
Re: Creating New Series using conditional
EViews Gareth wrote:You're declaring it as an alpha (which contains strings) and setting it equal to a number.
Either use series instead of alpha, or change the values to be strings (with the @str function).
Hi Gareth,
All done thanks!
- Wed Jun 21, 2017 11:09 am
- Forum: Data Manipulation
- Topic: Creating New Series using conditional
- Replies: 2
- Views: 3215
Creating New Series using conditional
Hi there, I would like to create a new series from an existing one called "airtemp". Im trying to get a (airtemp-24)^2 result when the value in "airtemp" exceeds 24; a (16-airtemp)^2 result when the value in "airtemp" is below 16 and "0" otherwise. But the fol...
- Tue Jun 20, 2017 1:34 pm
- Forum: Data Manipulation
- Topic: Descriptive Statistics by Date (hour)
- Replies: 5
- Views: 4951
Re: Descriptive Statistics by Date (hour)
Done!
Oh my days I feel so dumb now, haha.
Much appreciated.
Oh my days I feel so dumb now, haha.
Much appreciated.
- Tue Jun 20, 2017 1:33 pm
- Forum: Data Manipulation
- Topic: Descriptive Statistics by Date (hour)
- Replies: 5
- Views: 4951
Re: Descriptive Statistics by Date (hour)
EViews Gareth wrote:Depends on how the data are structured.
If you're in an hourly workfile, you could open the series and click on View->Descriptive Statistics->Stats by Classification and enter "@hour" as the classifying variable.
Thanks Gareth! I will start from there.
- Tue Jun 20, 2017 1:32 pm
- Forum: Data Manipulation
- Topic: Descriptive Statistics by Date (hour)
- Replies: 5
- Views: 4951
Re: Descriptive Statistics by Date (hour)
I am right now starting with the only solution I came up with. Im creating dummy variables depending of the "hour of the day" 1-24 in order to sample demand values using the "if" function with the dummy variable next to it.
But I believe Eviews provides more powerful short cuts.
But I believe Eviews provides more powerful short cuts.
- Tue Jun 20, 2017 1:27 pm
- Forum: Data Manipulation
- Topic: Descriptive Statistics by Date (hour)
- Replies: 5
- Views: 4951
Descriptive Statistics by Date (hour)
Hi all, I am experimenting for first time with Eviews in order to build a SARIMAX model for the UK Electricity Market Demand. Right now I am just testing the raw data (hourly demand from 2006.1.1 to 2016.12.31) I am now very much stuck since I have a time series of demand values for 10 years and nee...