|
|
A165438
|
|
Number a(n) of alternative sets of orthogonal contrasts available to partition variation between n levels of a categorical factor in analysis of variance, with each set described by a unique general linear model.
|
|
0
|
|
|
1, 3, 4, 8, 15, 34, 69, 152, 332, 751, 1698, 3905, 9020, 21051, 49356, 116505, 276217, 658091, 1573835, 3778152, 9098915, 21980209, 53241777, 129294912, 314714273, 767700735, 1876437054, 4595005570, 11271747564, 27695048780
(list;
graph;
refs;
listen;
history;
text;
internal format)
|
|
|
OFFSET
|
3,2
|
|
COMMENTS
|
Each set has n-1 orthogonal contrasts.
|
|
REFERENCES
|
Doncaster, C. P. & Davey, A. J. H. (2007) Analysis of Variance and Covariance: How to Choose and Construct Models for the Life Sciences. Cambridge: Cambridge University Press.
|
|
LINKS
|
|
|
FORMULA
|
For n=5,6,7: a(n) = -mod(n,2)*a([n-mod{n,2}]/2) + sum_{k=3..n-1} a[k]
For n>7: a(n) = -mod(n,2)*a([n-mod{n,2}]/2) + 2*a(n-1) + b(n) - b(n-1)
where b(m) = 0^mod(log(m,2),1) + mod(m-1,2)*0.5*a([m-mod{m,2}]/2)*(a[{m-mod(m,2)}/2]-1)
+ sum_{k=3..(m-1-mod[m-1,2])/2} a(m-k)*(a[k]-1)
|
|
EXAMPLE
|
A factor 'A' with n = 5 levels, has a(5) = 4 alternative sets of orthogonal
contrasts, each with n - 1 = 4 contrasts. The corresponding alternative
general linear models describing contrasts 'B', 'C', 'D', 'E' are:
B + C(B) + D(B) + E(D B)
B + C(B) + D(C B) + E(D C B)
B + C(B) + D(C B) + E(C B)
B + C(B) + D(B) + E(B)
|
|
CROSSREFS
|
|
|
KEYWORD
|
nonn
|
|
AUTHOR
|
C. Patrick Doncaster (cpd(AT)soton.ac.uk), Sep 18 2009
|
|
EXTENSIONS
|
Corrected and edited by C. P. Doncaster (cpd(AT)soton.ac.uk), Mar 02 2010
|
|
STATUS
|
approved
|
|
|
|