Removing confidence intervals in forecast package

noetsi

No cake for spunky
#1
I am trying to remove confidence intervals in my forecast because I don't use them and its a pain to remove them in excel. Does anyone know code to tell the forecast package not to show them? I have not found a way on line
 

Dason

Ambassador to the humans
#2
Just subset out the forecast directly
Code:
> library(forecast)
> fc <- forecast(gas)
> fc$mean
          Jan      Feb      Mar      Apr      May      Jun      Jul      Aug
1995                                                                        
1996 40460.93 40984.57 45295.32 46698.04 55425.32 59777.99 64328.11 61464.62
1997 40637.32 41159.60 45484.82 46889.42 55647.84 60013.09 64575.96 61696.62
          Sep      Oct      Nov      Dec
1995 54227.29 50915.06 46620.37 42982.42
1996 54484.03 51151.19 46832.16 43173.70
1997
 

hlsmith

Less is more. Stay pure. Stay poor.
#3
I never thought I would see a perfect interchange of info between you two. Usually @noetsi is too vague and leaves something out and @Dason complains or intentionally answers a different question. But above was so clean and concise. I think you all may have eventually mind melded!!!
 

noetsi

No cake for spunky
#4
I never thought I would see a perfect interchange of info between you two. Usually @noetsi is too vague and leaves something out and @Dason complains or intentionally answers a different question. But above was so clean and concise. I think you all may have eventually mind melded!!!
lol well I slowly learn

It is good to know Dason sometimes is answering different questions. That really confused me...
 

noetsi

No cake for spunky
#5
Just subset out the forecast directly
Code:
> library(forecast)
> fc <- forecast(gas)
> fc$mean
          Jan      Feb      Mar      Apr      May      Jun      Jul      Aug
1995                                                                       
1996 40460.93 40984.57 45295.32 46698.04 55425.32 59777.99 64328.11 61464.62
1997 40637.32 41159.60 45484.82 46889.42 55647.84 60013.09 64575.96 61696.62
          Sep      Oct      Nov      Dec
1995 54227.29 50915.06 46620.37 42982.42
1996 54484.03 51151.19 46832.16 43173.70
1997
In this case $mean is the column name you are interested in right (I don't know that field in forecast).
 

hlsmith

Less is more. Stay pure. Stay poor.
#6
Seems so, but I haven't used that package in 5 years. I would think that the intervals may then be something like $ucl or $upl or derived based on a SE. So fc$mean could get the estimate per @Dason.
 

Dason

Ambassador to the humans
#7
Look at the help page for your function. It documents the output.

Code:
?forecast.lm
Alternatively just use str to get a better understanding of the structure of the output directly.
 

noetsi

No cake for spunky
#8
I usually look for my questions on the internet rather than go into the documentation because I figure I will get lost there :)

We will now have R on our SQL server rather than have to use a separate connection. This was, we were told anyway, impossible to do on the old server, but we are moving to the 2019 SQL server.

Until now I am the only one who used R, but we might eventually get rid of our SAS and move to the new server which is free of course (and SAS is really hard to deal with administratively in terms of licenses - it is an immense pain dealing with their sales office. In the past it took absurd amounts of time to renew them or even find out how to renew them).

Amusingly I had to fight in the past to get R allowed at my agency (amusing because everyone here knows how enthusiastic I am about R). :p