da2m — transformation of daily data into monthly ones
ts=da2m(ts,ind,dropna)
ts = a daily timeseries
ind =
* -1 if the monthly data is to be the sum of the corresponding days
* 0 if the monthly data is to be the mean of the corresponding days
* n between 1 and 28 if the monthly data is to be the value of the n-th day of the month
* n > 29 if the monthly data is to be the value of the last day of the month
dropna = 'dropna' if the user wants to withdraw all NA values from the calculations (case ind=0 or -1) or to take the last non NA value of the quarter (case ind > 90)
if not entered then all NA values are considered
ts = the corresponding monthly time series
Let tsd be a daily ts from 2000/02/15 to 20007/08/18 with all week-end values set to %nan then if you run: 1)tsm=da2m(tsd,0) 2)tsm=da2m(tsd,0,'dropna') 3)tsm=da2m(tsd,1) 4)tsm=da2m(tsd,1,'dropna') 5)tsm=da2m(tsd,31) 6)tsm=da2m(tsd,31,'dropna') you obtain a monthly series - in case 1): from 2002m3 to 2007m7 with only NA values - in case 2): from 2002m3 to 2007m7 with values equal for each month to the mean of all non NA daily values in the month - in case 3): from 2002m3 to 2007m8 with values equal to the values of each first day in the month, which can be eventually NA - in case 4): from 2002m3 to 2007m8 with values equal to the values of each first day in the month ('dropna' does not change anything on this case) - in case 5): from 2002m2 to 2007m7 with values equal to the values of each last day in the month - in case 6): from 2002m2 to 2007m7 with values equal to last non NA value in the month