r - Determining when to log-transform (or perform other transformations) on a time-series and apply automatically -

Is there a series to check in any way or should it change in some other way?

I is a code in which I use to run many different data through the prediction, some data definitely need to be changed, though nothing does. As the code has been written to fully automate, it will be used by non-statisticians within the company so that they do not know that they should change the code to change the data based on the series. So I need tests that will check it for them and implement the changes accordingly.

Here is an example set of data that you can use:

  M & lt; - Matrix (C ("08Q1", "08Q2", "08Q3", "08Q4", "09Q1", "09Q2", "09Q3", "09Q4", "10Q1", "10Q2", "10Q3", " 10Q4 "11Q1", "11Q2", "11Q3", "11Q4", "12Q1", "12Q2", "12Q3", "12Q4", "13Q1", "13Q2", "13Q3", "13Q4" "14Q1", "14Q2", "14Q3", 5403.676, 6773.505, 7231.117, 7835.552, 5236.710, 5526.619, 6555.782, 11464.727, 7210.069, 7501.610, 8670.903, 10872.935, 820.923, 8153.393, 10196.448, 13244.502, 8356.733, 10188.442 , 10601.322, 12617.821, 11786.526, 10044.987, 11006.005, 15101.946, 10 992.273, 11421.189, 10731.312), ncol = 2, byrow = FALSE) New Mexico [, Length (m [1]]]  

I have found lambda for change from boxcoxfit () package geoR . Does anyone know how accurate it is to change the data is?

  ML < - boxcoxfit (Nu) Fit parameter: Lambda beta sigmasack 0.5 9 375.43 3649.39 N & T (- (($ nm $ lambda)) - 1) / mL $ lambda