###################################################################
# A list of possible future extensions of the ReGenesees package. #
#                                                                 #
# NOTE: This list is not necessarily up-to-date, and topics are   #
#       listed in no particular order.                            #
###################################################################


TOPIC --------------------------------------------------------------------------
     Bring Kott DAGJK into ReGenesees, by integrating package EVER into
     ReGenesees
COMPLEXITY/TIME
     High
MAIN DRIVER
     Me


TOPIC --------------------------------------------------------------------------
     Enable variance estimation of poverty and inequality estimators (e.g. Gini,
     Laeken indicators), via Deville influence functions approach
COMPLEXITY/TIME
     High
MAIN DRIVER
     Istat EU-SILC team, me


TOPIC --------------------------------------------------------------------------
     Enable exact variance estimation under multi-step calibration, using
     super residuals
COMPLEXITY/TIME
     High
MAIN DRIVER
     Me


TOPIC --------------------------------------------------------------------------
     Enable ReGenesses to digest calibration weights which have been computed
     externally (e.g. by other software), so as to provide correct variance
     estimates
COMPLEXITY/TIME
     Low
MAIN DRIVER
     ONS, some Istat social survey teams, me
STATUS
     *Done* - function 'ext.calibrated' released with ReGenesees 1.8


TOPIC --------------------------------------------------------------------------
     Enable ReGenesses to calibrate while simultaneously fulfilling range
     restrictions both on g-weights and on calibration weights (this would be
     far better than a heuristic post-calibration trimming procedure)
COMPLEXITY/TIME
     Medium
MAIN DRIVER
     ONS, some Istat survey teams (e.g. EUSILC), me
STATUS
     *Done* - function 'trimcal' released with ReGenesees 1.9


TOPIC --------------------------------------------------------------------------
     Devise a way to enable ReGenesees to automagically fill a population
     totals template by using a data mart of partially aggregated data as
     'universe' (instead of a complete microdata file)
COMPLEXITY/TIME
     ?
MAIN DRIVER
     All Istat household survey teams (as we do not have a population register
     yet), me


TOPIC --------------------------------------------------------------------------
     Integrate ReGenesees "partitioned" calibration approach with dedicated
     matrix algebra methods (e.g. multiplication, generalized inverse, QR) for
     sparse matrices
COMPLEXITY/TIME
     Medium
MAIN DRIVER
     Me, Istat team in charge of processing long-form Population Census sample


TOPIC --------------------------------------------------------------------------
     Parallelize variance estimation algorithms (mainly for domain estimation)
COMPLEXITY/TIME
     Medium
MAIN DRIVER
     Me

