Class to generate a set of trials for clustering. More...
#include <trial.h>
Collaboration diagram for trial_generator:
Public Member Functions | |
| trial_generator (size_t _max_k, size_t _max_reps, size_t _init_size, size_t _num_objects) | |
| Constructor to generate trials from 1 to max_k. | |
| trial_generator (size_t min_k, size_t _max_k, size_t _max_reps, size_t _init_size, size_t _num_objects) | |
| Constructor to generate trials from min_k to max_k. | |
| size_t | count () const |
| return iterations so far. | |
| bool | has_next () const |
| whether there are trials remaining. | |
| trial | next () |
| return parameters for next trial | |
| void | reset () |
| return to initial state | |
| size_t | num_trials () |
| Return total number of trials this will generate. | |
Public Attributes | |
| const size_t | max_k |
| maximum k to try | |
| const size_t | max_reps |
| max number of repetitions per k | |
| const size_t | init_size |
| initial size for samples before factoring in k, as per CLARA paper. | |
| const size_t | num_objects |
| number of elements in the data set; determines maximum sample size. | |
Detailed Description
Class to generate a set of trials for clustering.
This packages up state for the main trial loop and allows work to be farmed out to worker processes.
Definition at line 70 of file trial.h.
Constructor & Destructor Documentation
| trial_generator | ( | size_t | _max_k, |
| size_t | _max_reps, | ||
| size_t | _init_size, | ||
| size_t | _num_objects | ||
| ) |
| trial_generator | ( | size_t | min_k, |
| size_t | _max_k, | ||
| size_t | _max_reps, | ||
| size_t | _init_size, | ||
| size_t | _num_objects | ||
| ) |
Member Function Documentation
| bool has_next | ( | ) | const |
| size_t num_trials | ( | ) |
| void reset | ( | ) |
return to initial state
Member Data Documentation
| const size_t init_size |
| const size_t num_objects |
The documentation for this class was generated from the following files:
