Code Release

New code along with a ton of data files fro 2007-2009 have been released. The following utilities are present in this release

1. screen: A utility to screen parameters for a particular value has been created.
2. concat: A utility to concatenate a data set across multiple years
3. merge: a utility to merge two data-sets and align the data by SEQN
4. correl: A utility to calculate a correlation coefficient between two variables in the same file
5. correlall: a utility to return the correlation coefficient between all variables in the same file
6. correlall-findstrong: the same as above but it only returns coefficients for a set of variabels with a high n and high r-squared value.

example:
concat glu_e.txt glu_f.txt > glu.txt
concat bmx_e.txt bmx_f.txt > bmx.txt
merge glu.txt bmx.txt > glubmx.txt
correlall-findstrong glubmx.txt 0.3

The above example will find all correlations between the fasting glucose and biometric values. The results of this are as follows:

BMXWT LBXIN 0.360892 6322
BMXBMI LBXIN 0.405838 6314
BMXARMC LBXIN 0.3517 6206
BMXWAIST LBXIN 0.387872 6154
BMXWT LBDINSI 0.360892 6322
BMXBMI LBDINSI 0.405838 6314
BMXARMC LBDINSI 0.3517 6206
BMXWAIST LBDINSI 0.387872 6154

It returns the two compared variables, there correlation coefficient and the number of values compared.