2006.08.05	Arrange fast "glm"-like for use in gwaa.scan

2006.08.17	implement a function to read WTCCC type of data
		(directly from R or at least indirectly through a
		companion parogram)

2006.10.11	optimisation of emp.ccfast and emp.qtscore 
		As most of the time in C is spent on extracting data 
		from 2-bit format, most efficient would be to pass list 
		of permuted indexes to the C procedure.

2006.10.11	check.marker test for redundancy: also limit by region (low priority)

2006.10.25	additional QC for markers: check for average heterozygosity
		(there is heteroz procedure, but it should be re-implemented in C)
		within a person (can pick up contamination), scatter-plot
		of IBS/var(IBS), as in GRR, to pick up related people.

2006.11.05	For check.marker: also test missingness by person, not 
		only marker

2006.11.06	Operations with formula and data by qtscore() and scan.glm() should be 
		more in line with what is done by generic glm() and lm()

2006.11.06	Arrange formal class for check.marker-class

2006.11.06	Arrange run.haplo.score formula in accordance with 
		qtscore

2006.12.21	Fix ccfast so it will allow for correct analysis of X chromosome

2006.12.21	Analysis of X-chromosome data is totally wrong for emp.qtscore, 
		ccfast, emp.ccfast
		
