Weighting and Estimation for Panel Data with Non-Response
Generate test data
Weighting and Estimation for Panel Data with Non-Response
Calibration weighting and estimation
Calibration weighting and variance estimation for panel data
Create or modify a CalSSB object
Create a CalSSB object from a ReGenesees/cal.analytic object
Crossing several factor variables
Creation of linear combination matrices
MatchVarNames
OrderedVarNames
Variance estimation for panel data
Print method for calSSB
Print method for calSSBwide
Rearrange output from CalibrateSSB (calSSB object). Ready for input to...
Functions to calculate weights, estimates of changes and corresponding variance estimates for panel data with non-response. Partially overlapping samples are handled. Initially, weights are calculated by linear calibration. By default, the survey package is used for this purpose. It is also possible to use ReGenesees, which can be installed from <https://github.com/DiegoZardetto/ReGenesees>. Variances of linear combinations (changes and averages) and ratios are calculated from a covariance matrix based on residuals according to the calibration model. The methodology was presented at the conference, The Use of R in Official Statistics, and is described in Langsrud (2016) <http://www.revistadestatistica.ro/wp-content/uploads/2016/06/RRS2_2016_A021.pdf>.
Useful links