🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
Calcite | Level 5

## Predicting new observation

Good day,

I would like to predict new variables based on the precious values. the data below shows the variables.

 Month Average volume for month September,16 10510 October,16 13288 November,16 15841 December,16 17535 January,17 19415 February,17 21221 March,17 23321 April,17 25314 May,17 27240 June,17 29430 July,17 31300 August,17 33090 September,17 34983 October,17 37065 November,17 December,17 January,18 February,18 March,18 April,18 May,18

How can i best use the values from previous volumes to predict volumes in missing months.

thank you

1 ACCEPTED SOLUTION

Accepted Solutions
SAS Employee

## Re: Predicting new observation Using SAS Statistics

Kasharp provided above a possible way of generating predictions using PROC ESM which supports exponential smoothing models.

Since you say you use SAS Studio, I suggest you look at the supplied Forecasting tasks on the left panel.  This short video gives you an introductory tutorial.

http://video.sas.com/detail/videos/how-to-tutorials/video/4414522302001/time-series-modeling-and-for...

6 REPLIES 6
Diamond | Level 26

## Re: Predicting new observation

What is it your asking?  Its your data, so I assume you would know how you want to derive future values?  There are many possiblities, such as taking the last value and carrying it forward, lowest value, max value, mean/median value, previous minus current carried forward etc.  As I have no idea what your trying to achieve i don't see how I  could suggest one.

Calcite | Level 5

## Re: Predicting new observation

i have sales volumes for the months stated. what i want is to predict future sales volumes for months without readings. in other ways h do i analyze the data above to draw patterns and predict the the future volumes

Super User

## Re: Predicting new observation

Suggest you to post it at Forecast forum,since it is about time series analysis.

``````data have;
infile cards expandtabs truncover dlm=' ';
input Month : anydtdte32.	Average;
format month monyy7.;
cards;
Sep16	10510
Oct16	13288
Nov16	15841
Dec16	17535
Jan17	19415
Feb17	21221
Mar17	23321
Apr17	25314
May17	27240
Jun17	29430
Jul17	31300
Aug17	33090
Sep17	34983
Oct17	37065
Nov17
Dec17
Jan18
Feb18
Mar18
Apr18
May18
;
run;
proc esm data=have out=nextyear ;
id month interval=month;
run;
proc sgplot data=nextyear;
series x=month y=average;
run;
``````
Calcite | Level 5

## Re: Predicting new observation

Thank you, your code has helped me clean and formart my data in preparation for Time series Forecasting

that informat to convert string values to dates . Thanx

Calcite | Level 5

## Predicting new observation Using SAS Statistics

Good day,

I would like to predict new variables based on the precious values. the data below shows the variables.

 Month volume for month September,16 10510 October,16 13288 November,16 15841 December,16 17535 January,17 19415 February,17 21221 March,17 23321 April,17 25314 May,17 27240 June,17 29430 July,17 31300 August,17 33090 September,17 34983 October,17 37065 November,17 December,17 January,18 February,18 March,18 April,18 May,18

How can i best use the values from previous volumes to predict volumes in missing months. Can this be aced using SAS Studio and how. Please assist

thank you

SAS Employee

## Re: Predicting new observation Using SAS Statistics

Kasharp provided above a possible way of generating predictions using PROC ESM which supports exponential smoothing models.

Since you say you use SAS Studio, I suggest you look at the supplied Forecasting tasks on the left panel.  This short video gives you an introductory tutorial.

http://video.sas.com/detail/videos/how-to-tutorials/video/4414522302001/time-series-modeling-and-for...

Discussion stats
• 6 replies
• 1367 views
• 1 like
• 4 in conversation