Bug: stats::lm() produces NA-containing ty.

[t-pollington]

Following Issue #4 this is a minimal bug correction to prevent a user running bcajack2() if their Y choice leads to B * pct approximately being less than or equal to n aka ncol(Y).

Otherwise lm() will produce some NA entries in ty. as there are insufficient observations (length(ip)) to solve the matrix equation in lm(), considering the number of dimensions of x in lm() is p aka ncol(Y). Happy to explain further, as I'm aware my explanation is spread across this PR and Issue #4.

[t-pollington]

Additional commits to bcajack() (note not bcajack2() this time) following 2 possible bugs:

• Code correction and some refactoring of the loop to get aa and ssj. Imat[Ij, ] seemed to be a mistake as the cols of Imat represent the m groups rather than its rows.

• Code refactoring. Since the line:
  u. <- 2 * t. - s.
  gave a warning of different t. & s. lengths when applied to my data. Code refactoring followed "Estimation and Accuracy After Model Selection" by Efron, p995 as noted in "The Automatic Construction..." paper.

[bnaras]

I went over these changed files and am finding it difficult to distinguish the substantive changes from the non-substantive ones (such as indentation, new lines etc.) Could you please keep the original indentation? Thank you.

[t-pollington]

Please bear with me. Making a few corrections to my code so could be a few weeks before re-submission.