Missing values are ommited in the regression. (Will need to check
this.)
WONTFIX (Michael): There is still some error associated with this
function, will need to investigate further. Will need a rule for
this, when the fluctuation is large and data are sufficient then
take the lsgr, otherwise the geogr.
In order to ensure the growth rate calculated is reliable, the
following rule are applied.
50% of the data must be present.
The length of the time series must be greater than n.
Otherwise the growth will not be computed.