merge ANOVA_exact2 with ANOVA_exact?

[tbates]

I wonder if the need for ANOVA_exact2 could be obviated by adding a bigsample= TRUE parameter to ANOVA_exact to cope with the difficulty back-filling small samples under the exact constraint?

do you mind suggestions?

Another would be for the error on illegal design strings to include a legal example. e.g.

Problem in the design argument: must consist of the number of levels (2-999) and factor-type (lower case b (between) or w (within)). e.g "2b" or "2b*3w"

instead of

Problem in the design argument: must input number of levels as integer (2-999) and factor-type (between or within) as lower case b (between) or w (within). e.g

[arcaldwell49]

Hi Tim, I don't have the capacity right now to handle that big of an update (and it might run cause some issues with a re-submission to CRAN).

However, I welcome pull requests.

[tbates]

I'll have a look at coding this up.