|
pulsatrix 1.1.0
|
How often the sets or intervals held the truth, and how large they were. More...
#include <conformal.hpp>
Public Attributes | |
| int64_t | n = 0 |
| double | coverage = 0.0 |
| The fraction that held the truth; compare with 1 − alpha. | |
| double | mean_size = 0.0 |
| The mean set size, or the mean interval width. | |
| int64_t | empty = 0 |
| Classification only: how many sets were empty. | |
| std::vector< int64_t > | size_counts |
| Classification only: size_counts[s] sets had s classes, and coverage_by_size[s] of them held the truth (0 where there were none). Undercoverage of the large sets is a sign the score doesn't adapt to hard inputs. | |
| std::vector< double > | coverage_by_size |
| std::vector< double > | misses |
| Per input: 1 where the truth was missed, 0 where it was held; PerSliceError() on it tests each slice's coverage. | |
How often the sets or intervals held the truth, and how large they were.
| double pulsatrix::CoverageReport::coverage = 0.0 |
The fraction that held the truth; compare with 1 − alpha.
| std::vector<double> pulsatrix::CoverageReport::coverage_by_size |
| int64_t pulsatrix::CoverageReport::empty = 0 |
Classification only: how many sets were empty.
| double pulsatrix::CoverageReport::mean_size = 0.0 |
The mean set size, or the mean interval width.
| std::vector<double> pulsatrix::CoverageReport::misses |
Per input: 1 where the truth was missed, 0 where it was held; PerSliceError() on it tests each slice's coverage.
| int64_t pulsatrix::CoverageReport::n = 0 |
| std::vector<int64_t> pulsatrix::CoverageReport::size_counts |
Classification only: size_counts[s] sets had s classes, and coverage_by_size[s] of them held the truth (0 where there were none). Undercoverage of the large sets is a sign the score doesn't adapt to hard inputs.