Changeset 748 for trunk/yat/statistics


Ignore:
Timestamp:
Feb 11, 2007, 6:48:22 PM (15 years ago)
Author:
Peter
Message:

documentation for SNR improved. fixes #123

File:
1 edited

Legend:

Unmodified
Added
Removed
  • trunk/yat/statistics/SNR.h

    r703 r748  
    4545     sometimes referred to as Golub score. The score is the ratio
    4646     between difference in mean and the sum of standard deviations
    47      for two groups.  \f$ \frac{ \frac{1}{n_x}\sum x_i -
    48      \frac{1}{n_y}\sum y_i }{ \sigma_x + \sigma_y} \f$ where \f$
    49      \sigma \f$ is standard deviation.
     47     for two groups:  \f$ \frac{ m_x-m_y}{ s_x + s_y} \f$ where \f$
     48     s \f$ is standard deviation.
    5049  */
    5150  class SNR : public Score
     
    5857    SNR(bool absolute=true);
    5958
    60     ///
    61     /// Calculates the value of SNR score. @return SNR score if
    62     /// absolute=true absolute value of t-score is returned
    63     ///
     59    /**
     60       SNR is defined as \f$ \frac{m_x-m_y}{s_x+s_y} \f$ where \f$ m
     61       \f$ and \f$ s \f$ are mean and standard deviation,
     62       respectively. @see Averager
     63
     64       @return SNR score. If absolute=true absolute value of SNR is
     65       returned
     66    */
    6467    double score(const classifier::Target& target,
    6568                 const utility::vector& value);
    6669
    67     ///
    68     /// Weighted version of SNR @return t-score if absolute=true
    69     /// absolute value of t-score is returned.
    70     ///
     70    /**
     71       SNR is defined as \f$ \frac{m_x-m_y}{s_x+s_y} \f$ where \f$ m
     72       \f$ and \f$ s \f$ are weighted versions of mean and standard
     73       deviation, respectively. @see AveragerWeighted
     74
     75       @return SNR score. If absolute=true absolute value of SNR is
     76       returned
     77    */
    7178    double score(const classifier::Target& target,
    7279                 const classifier::DataLookupWeighted1D& value);         
    7380
    74     ///
    75     /// Weighted version of SNR @return t-score if absolute=true
    76     /// absolute value of t-score is returned.
    77     ///
     81    /**
     82       SNR is defined as \f$ \frac{m_x-m_y}{s_x+s_y} \f$ where \f$ m
     83       \f$ and \f$ s \f$ are weighted versions of mean and standard
     84       deviation, respectively. @see AveragerWeighted
     85
     86       @return SNR score. If absolute=true absolute value of SNR is
     87       returned
     88    */
    7889    double score(const classifier::Target& target,
    7990                 const utility::vector& value,
Note: See TracChangeset for help on using the changeset viewer.