Simulate - lucadellasantina/ObjectFinder GitHub Wiki

ObjectFinder can simulate the random assignment of specific statistics using a monte-carlo simulator approach:

Simulation Actions:

Simulation Settings:

Set of test objects: Objects to be used as reference for the simulation. Default: current objects
Number of objects to process: Total number of objects to include in the simulation. Default: all valid objects in the set.
Number of iterations: Number of times objects statistic is randomly assigned. Default: 2000 times
Tested parameter: Parameter of test objects being randomized by the simulator.
Computed statistic: Readout statistic evaluated in the randomized set.
Type of second test set: If selected parameter involves a second test set, specify the nature of it. Default: Objects.
Second set of test objects: Either another objects set or a binary mask image.

Completed simulations:

Complete simulations: list of simulations performed on the selected set of objects, select one item from the list to enable actions (delete, test hypothesis,...) on it.
Rename simulation: changes the name of selected simulation from the one automatically assigned
Copy simulation: duplicates the selected simulation
Delete simulation: removes the selected simulation from disk (This option cannot be undone).
Clear simulation list: removes all simulations from disk (This option cannot be undone).

Hypothesis testing:

Tests the likelihood that the empirical set of objects is drawn from randomized simulation.

Null Hypothesis: This is the null hypothesis being tested by the randomization test
Significance level: P-value below which the null hypothesis is rejected
Two tailed test: Tests whether the null hypothesis is to be rehected on both ends of the randomized distribution

Technical details:

ObjectFinder uses a Mersenne Twisterr as default pseudorandom number generator for the Monte-Carlo simulations