add a sample dataset to explore the functions without a datafile.

add documentation for sample dataset

The accès to various various is global which clutters the R environment. Cleanup access control.

return the function result as user defined name to avoid conflict/overwriting in sample data objects
cc_define_file() returns an object cc_dataset
and sample dataset is also called cc_dataset

data set is loaded but it does not show up at runtime, throws an error

y standing alone is giving problems. I think it is better to keep it within data and use the i index. please refer comments in #12

use this convention: (only for package, not for internal functions)
this is only for developer

prefix all exported functions with cc: ccread, ccmerge, ccsplit, ccplot, ccout

add ability to read data from excel sheets

Add a check:
2. If all pairs have p value significant than user defined alpha_merge, go to bonferroni

Currently, the pairwise.t.test collects max p value for both ordinal and nominal. This is wrong.

for nominal it should be max of the whole p matrix

for ordinal, it should be the max of the diagonal of the whole p matrix

Sri,
Would you please push the merge.R with the merging function you made?

add documentation for each function

Add two checks for merge.step

1. If there are only two levels in a factor predictor, do not split the predictor