as_sts() for IncidenceMatri should have correct start date

[nickreich]

I think the default time is getting passed to it as a start time, see the following example:

> dat <- read.csv('data/province-biweek-counts.csv')
> inc <- IncidenceMatrix$new(1+reshape2::acast(dat,province~date_sick,value.var='cases'))
> inc$as_sts()
-- An object of class sts -- 
freq:		 52 
start:		 2000 1 
dim(observed):	 208 77
...

note the start of 2000 1 which is the default for sts objects. As this is explicitly not the start of the timeseries:

> reshape2::acast(dat,province~date_sick,value.var='cases')
   2006-01-01 2006-01-15 2006-01-29 2006-02-12 2006-02-26 2006-03-12 2006-03-26 2006-04-09 2006-04-23
10        265        162        211        113        165        146        137        117         74
11         27         30         37         37         35         32         29         32         24