# Template talk:PAGR

## Precision parameter

Jimp, do you suppose you could add a precision parameter to this template just as you did for me previously at Template:PGR? It would be preferred that the parameter be called `prec`. Cheers, Hwy43 (talk) 06:34, 23 June 2014 (UTC)

I've added the parameter. I also changed the template it calls from {{decimals}} to {{rnd}}. This means that you get a true minus sign rather than a hyphen. Jimp 12:45, 23 June 2014 (UTC)
Thanks! Hwy43 (talk) 19:01, 23 June 2014 (UTC)

## How to use this template with different references dates

Hi, Is it possible to use this template when the references dates of various census years differ? More specifically: This template is used on Template:Philippine Census. In the current situation it does not take into account that the census of the year 1995 was taken on September 1st and the census of 2007 was on August 1st, while the census of other years was on May 1st. The means that the per annum growth rates of the last four periods are slightly incorrect. I hope somebody can correct this, or tell me how do it. Magalhães (talk) 14:09, 31 July 2014 (UTC)

I suppose the best way is to use days. Because both years are leap years, September 1 is the 244th day of the year (244/365=0.668) so use 1995.668, August 1 the 213th day so 2007.584, and May 1 the 121st day so YYYY.332. For example {{`PAGR|1980.668|110043|1990.332|180288`}} → "5.24%". 18:01, 31 July 2014 (UTC)
Thanks, Magalhães (talk) 05:43, 2 August 2014 (UTC)
My two penn'orth:
If you're going to be as precise as days, then I suggest you keep it in days rather than decimals of year. Excel on my laptop is perfectly capable of working out daily interest on a mortgage, by deriving the 365th (or 366th) root of the nominal yearly rate. Even though it can happily produce 10D, in fact 6D is precise enough. Once you have derived a daily growth rate by formula, let's call it PDGR, then you only need to take its 365th (or 366th) power to get PAGR. Don't know how you decide - maybe 365.25th power??
Of course, months might be as close as you need. Either way, I think the fraction part of the input should be days (months), not decimal fraction of a year. SO 1 sep 1980 => 1980.244 and 1 may 1990 => 1990.121 (or 1980.09 and 1990.05)
Problem with all this precision is that it doesn't compute on display, so although the true PAGR may be shown, taking into account month, what displays is seemingly incorrect.
Historical populations
Year Pop. ±% p.a.
1990 56,363 —
1995 66,683 +3.42%
2000 77,700 +3.11%
2007 92,297 +2.49%
2010 96,792 +1.60%
Population census of Tagbilaran
Year Pop. ±% p.a.
1990 56,363 —
1995 66,683 +3.20%
2000 77,700 +3.33%
2007 92,297 +2.40%
2010 96,792 +1.75%

For instance pop census growth to 2010 is shown as 1.75%, but if you grow 2007 number by that amount, it comes out as 97228. Maybe template needs to show month of year too. 112.198.82.1 (talk) 14:52, 11 October 2014 (UTC)

Forget that – anything is messy. What I have done instead is add a further footnote to the table, explaining discrepancies. See Tagbilaran#demographics 112.198.82.163 (talk) 04:10, 13 October 2014 (UTC)