results {mspath} | R Documentation |
Get results of previous calculations.
results(calc)
calc |
A previously evaluated calculator, which should be a
mspathCalculator or subclass. |
Currently numeric
of length 6
with the following entries, each being a sum over the active cases:
1: -2 Log-likelihood |
Minus twice the log-likelihood. This will
only be a
meaningful value if the calculation had an appropriate
do.what . |
2: Number of Cases |
This counts subjects, each of which may have several records. |
3: Good Paths |
Number of distinct paths consistent with the data and the model. |
4: Good Nodes |
Presumed to be the best single predictor of effort, in particular of floating-point operations. This is the number of unique nodes on good paths. |
5: Bad Nodes |
Number of nodes considered that fell outside of the good paths. May vary with the pruning strategy used, but at any rate note that these nodes need not be constructed. |
6: Good Path Nodes |
This is the sum of all the nodes in all the good paths, ignoring the fact that some nodes are shared between paths. This is a measure of the work effort that a naive implementation would require. The current implementation is not naive. |
Which of those values has useful data depends in particular on the
value of do.what
. Other, more complex, types of return values
are possible.
Ross Boylan
mspathCalculator, estimateWork
.