James Marett 3/29/2015

Like

1131 Views

Taylor: Which function works best with real world problem of knowing the next five years electrical generation plans and capacities and then knowing (hoping) that a value will be reached some 10 years after the last data point:

2012: 100 2014: 150 2016: 180 2030: 300

Which function will smooth the line from 2016 to 2030.... and still allow the factual data inputs for the first three parameters.

Step function "steps" but no growth curve smoothing between data points?

Self Answer: Correct approach is to use "Interp" function: Interp(2012, 100, 2014, 150, 2016, 180, 2030, 300)

Smooth and Smooth Growth seem to work if only 2 or 3 data year entries are made, allowing the program to compute polynomial regression between points. If too many points are entered (or too many outliers) program will diagnose error.

Interp seems to work well with real world electrical generation capacities increases over multiple years for the near term.... and a "guess" projection several years in the future.

James

Taylor Binnington 3/31/2015
 

Like

1127 Views

Hi James,

I usually use the Interp() function as a 'Swiss army knife) of time-series functions: my rationale is that if I don't know how a value will be approached, then I might as well assume it is approached linearly.

It sounds like you've experimented a little bit with the time series functions. Note that you can view a list of them (and their associated descriptions) by selecting the 'Function' option from the context menu (the orange "E") in the expression field. Select Function Type: Time-Series, and you will be shown a list of functions.

Hope this helps,

Taylor