Public Types | |
| enum | Type { unknown = -1, min = 0, max, range, nmin, nmax, mean, median, sum, stdDev, variance, rms, stdDevMAD, minority, majority, pdens, pcount, quantile, shannonEntropy, quadraticEntropy, center, rank, posOpenness, negOpenness, pdist, null, Count } |
| Enumeration defining different cell feature types (used by opalsCell) More... | |
| enum | Mode { undefined = 0, accumulative, selective, selectiveUpper = selective, selectiveLower } |
Public Member Functions | |
| StatFeature (Type type, Mode m=Mode::undefined) | |
| depending on type, m is ignored. | |
| StatFeature (Type type, unsigned n) | |
| only for nmin and nmax | |
| StatFeature (Type type, double pOrBinWidth, Mode m=Mode::accumulative) | |
| only for quantile, minority, majority, shannonEntropy, quadraticEntropy. m is ignored except for Type::quantile | |
| Type | featureType () const |
| unsigned | n () const |
| double | p () const |
| double | binWidth () const |
| Mode | mode () const |
| void | featureType (const Type &Ire_featureType) |
| void | n (const unsigned &Iru_n) |
| void | p (const double &Ird_p) |
| void | binWidth (const double &Ird_binWidth) |
| void | mode (const Mode &m) |
| unsigned | minCount () const |
| returns minimum number of elements needed to be computable | |
| enum Type |
Enumeration defining different cell feature types (used by opalsCell)
Used for indexing. Hence unscoped, with implicit conversion to int.
| Enumerator | |
|---|---|
| min | lowest |
| max | highest |
| range | difference between highest and lowest |
| nmin | n lowest |
| nmax | n highest |
| mean | mean |
| median | median |
| sum | sum |
| stdDev | standard deviation \(\sqrt{\sum_{i=1}^{N}(r_i - \bar{r})^2 / (N-1) }\) |
| variance | (sample) variance \(\sum_{i=1}^{N}(r_i - \bar{r})^2 / (N-1)\) |
| rms | root mean square \(\sqrt{\sum_{i=1}^{N} r_i^2 / N}\) |
| stdDevMAD | robust standard deviation estimator computed from the median of absolute deviations from the median \(1.4826 \cdot median_i(\left|r_i - median_j(r_j)\right|)\). In literature this robust estimator is often referred to as normalized MAD (NMAD) |
| minority | the class with the lowest relative frequency (histogram based) |
| majority | the class with the highest relative frequency (histogram based) |
| pdens | point density |
| pcount | point count |
| quantile | p-quantile (p=0..1) |
| shannonEntropy | shannon entropy, measure of information content |
| quadraticEntropy | quadratic entropy, measure of information content |
| center | select closest to cell center |
| rank | quantile rank [0..100] |
| posOpenness | positive openness |
| negOpenness | negative openness |
| pdist | average linear point distance (see DM::ILinearDistance2D for details) |
| null | representing void feature |
| Count | number of available StatFeatures |