📄 stat_utils.h
字号:
/*------------------------------------------------------------------------------*
* File Name: Stat_Utils.h *
* Creation: GJL 4/1/2003 *
* Purpose: Header file for Stat_Utils.c. *
* Copyright (c) OriginLab Corp. 2003-2007 *
* All Rights Reserved *
* *
* Modification Log: *
*------------------------------------------------------------------------------*/
#ifndef _STAT_UTILS_H
#define _STAT_UTILS_H
////////////////////////////////////////////////////////////////////////////////////
// Included header files
//////////////////////////////////////////////////////////////////////////////////
//
// System includes
//#include <origin.h> // Most Origin related classes
//#include <Tree.h> // Tree class
//#include <NAG\nag_types.h> // NAG structures
//#include <NAG\OCN_g01.h> // NAG Basic Statistics functions
////////////////////////////////////////////////////////////////////////////////////
// Define onstants and non-localized strings
////////////////////////////////////////////////////////////////////////////////////
//
#define SU_MIN_NPTS 2
#define SU_DEFAULT_CONF_LEVEL 0.95
#define SU_DEFAULT_TOLERANCE 0.000001
#define SU_DELIM '|'
#define SU_P1 "P1"
#define SU_PN "P%d"
#define SU_PARAMS_NODE "Parameters"
#define LR_COL_NAMES "Intercept|Slope"
#define PR_P0_NAME "A"
#define PR_PN_NAME "B%d"
#define PR_N_OUTPUT_COLS 6
#define PR_FITX_COL 0
#define PR_FITY_COL 1
#define PR_LCB_COL 2
#define PR_UCB_COL 3
#define PR_LPB_COL 4
#define PR_UPB_COL 5
////////////////////////////////////////////////////////////////////////////////////
// SumStats structure
////////////////////////////////////////////////////////////////////////////////////
//
typedef struct tagSumStats
{
int N;
double dSum;
double dWeightSum;
double dMean;
double dSD;
double dKurtosis;
double dSkewness;
double dSE;
double dVariance;
double dCoefVar;
double dCIL;
double dCIU;
double dMin;
double dMax;
double dRange;
int nMissing;
} SumStats;
////////////////////////////////////////////////////////////////////////////////////
// Enumerations
////////////////////////////////////////////////////////////////////////////////////
//
enum {FITLR_X, FITLR_Y, FITLR_LCB, FITLR_UCB, FITLR_LPB, FITLR_UPB, FITLR_NUM_COLS};
////////////////////////////////////////////////////////////////////////////////////
// Function prototypes
////////////////////////////////////////////////////////////////////////////////////
//
/** >Statistics
Function to compute summary statistics for a vector passing results back in a tree.
*/
bool stat_descriptive(const vector& vData, TreeNode& trWhole, int nOffset = 0, const vector* pvWeight = NULL);
/** >Statistics
Function to compute summary statistics for a vector passing results back in a SumStats structure.
*/
bool stat_summary(SumStats& ss, const vector& vData, const vector* pvWeight = NULL,
double dConfLevel = SU_DEFAULT_CONF_LEVEL);
/** >Statistics
Function to compute percentiles for a vector.
*/
bool stat_percentiles(vector& vPercentiles, const vector& vData, const vector& vPercents, bool bInterpolate = false);
/** >Statistics
Function to perform multiple linear regression using the NAG function nag_regsn_mult_linear.
Parameters:
nPts=Input number of rows (or data points) in the input matrix
nTdx=Input number of columns (or independent variables) in the input matrix
mX=Input matrix containing data points of the independent variables
vY=Input vector containing data points of dependent variable
vWT=Input vector containing weights of data points, NULL if weighting not used
dRss=Output residual sum of squares for the regression
dDf=Output degrees of freedom associated with dRss
vB=Output least-squares estimates of the parameters of the regression model
vSE=Output standard errors of the parameter estimates given in vB
vCOV=Output variance-covariance matrix of estimated parameters in vB
vRES=Output weighted residuals
vH=Output diagonal elements of H
mQ=Output results of the QR decomposition
bSvd=Output flag set TRUE if singular value decomposition was performed, otherwise FALSE
iRank=Output rank of the independent variables
vP=Output details of QR decomposition if SVD is used
vCOMAR=Output information which is needed by nag_regsn_mult_linear_newyvar_if bSvd is TRUE
dTol=Input tolerance used to decide rank of independent variables, default is MR_DEFAULT_TOLERANCE
bThroughZero=Input flag forcing regression line through origin, default is FALSE
Return:
Returns the results of the NAG nag_regsn_mult_linear function.
*/
int stat_multiple_linear_regression( int nPts, int nTdx, matrix& mX, vector& vY, vector& vWT,
double& dRss, double& dDf, vector& vB, vector& vSE, vector& vCOV, vector& vRES,
vector& vH, matrix& mQ, BOOL& bSvd, int& iRank, vector& vP, vector& vCOMAR,
double dTol = SU_DEFAULT_TOLERANCE, BOOL bThroughZero = FALSE );
/** >Statistics
Perform a linear regression on the input curve.
Parameters:
crvData=Input, the X and Y coordinates of the sample data
mResultCurves=Output, the result curves where
Col(0) is the X coordinates of the fit line
Col(1) is the Y coordinates of the fit line
Col(2) is the lower confidence band
Col(3) is the upper confidence band
Col(4) is the lower prediction band
Col(5) is the upper prediction band
trLR=Input settings & Output results of the linear regression
trLR.GUI.Fit.ThroughZero.nVal // Force fit to pass thru zero
trLR.GUI.Fit.FixSlope.nVal // force slope to be fixed
trLR.GUI.Fit.FixSlopeAt.dVal // fixed value of slope (Note: cannot fix slope and thru zero at same time, should return error from LLOC)
trLR.GUI.Fit.ErrBarWeight.nVal // Use error column for wt - use (1/err^2) as wt factor
trLR.GUI.Fit.UseReducedChiSq.nVal // Scale parameter errors with reduced chisqr
trLR.GUI.ResultCurves.Points.nVal // Number of points in fit curve
trLR.GUI.ResultCurves.ConfBands.nVal // Create confidence bands - if not set, then matrix will have empty columns
trLR.GUI.ResultCurves.PredBands.nVal // Create prediction bands - if not set, then matrix will have empty columns
trLR.GUI.ResultCurves.Confidence.dVal // Confidence value to be used
trLR.Calculation.Control.UseDataXRange.nVal // Option = 1 to use data range, = 0 to use X1, X2
trLR.Calculation.Control.X1.dVal // Default X minimum for fit curve
trLR.Calculation.Control.X2.dVal // Default X maximum for fit curve
trLR.Calculation.Parameters.P1.Name.strVal // Parameter Name...A
trLR.Calculation.Parameters.P1.Value.dVal // Parameter Value
trLR.Calculation.Parameters.P1.Error.dVal // Parameter Error
trLR.Calculation.Parameters.P1.Vary.nVal // Parameter Vary
trLR.Calculation.Parameters.P1.tValue.dVal // Parameter t-Value
trLR.Calculation.Parameters.P1.LCI.dVal // Parameter LCI
trLR.Calculation.Parameters.P1.UCI.dVal // Parameter UCI
trLR.Calculation.Parameters.P1.Prob.dVal // Parameter Prob > |t|
trLR.Calculation.Parameters.P2.Name.strVal // Parameter Name...B
trLR.Calculation.Parameters.P2.Value.dVal
trLR.Calculation.Parameters.P2.Error.dVal
trLR.Calculation.Parameters.P2.Vary.nVal
trLR.Calculation.Parameters.P2.tValue.dVal
trLR.Calculation.Parameters.P2.LCI.dVal
trLR.Calculation.Parameters.P2.UCI.dVal
trLR.Calculation.Parameters.P2.Prob.dVal
trLR.Calculation.Stat.Rvalue.dVal // R-value
trLR.Calculation.Stat.RSqCOD.dVal // RSq-COD value
trLR.Calculation.Stat.AdjRSq.dVal // Adjusted RSq
trLR.Calculation.Stat.RMSESD.dVal // RMSE-Standard Dev.
trLR.Calculation.Stat.N.nVal // No. of points in fit
trLR.Calculation.ANOVA.Model.DOF.nVal // Model DOF
trLR.Calculation.ANOVA.Model.SSq.dVal // Model SS
trLR.Calculation.ANOVA.Model.MeanSq.dVal // Model Mean Sq
trLR.Calculation.ANOVA.Error.DOF.nVal // Error DOF
trLR.Calculation.ANOVA.Error.SSq.dVal // Error SS
trLR.Calculation.ANOVA.Error.MeanSq.dVal // Error Mean Sq
trLR.Calculation.ANOVA.Total.DOF.dVal // Total degrees of freedom
trLR.Calculation.ANOVA.Total.SSq.dVal // Total SS
trLR.Calculation.ANOVA.Fvalue.dVal // F-value
trLR.Calculation.ANOVA.Pvalue.dVal // P-value
Return:
Returns ERROR_NO_ERROR on successful exit and an error code on failure.
NAG error codes: Negative 1 * NAG error code
ERROR_INVALID_CURVE: Invalid input curve
ERROR_INVALID_TREENODE: Invalid TreeNode
ERROR_TO_FEW_PTS: Not enough points in the curve
ERROR_X_RANGE: All the elements of X are equal
ERROR_SETTINGS: 'ThroughZero' and 'FixSlope' can not be both enabled
*/
int stat_linear_fit(const curvebase& crvData, matrix& mResultCurves, TreeNode& trLR);
/** >Statistics
Perform a multiple linear regression using the NAG function nag_regsn_mult_linear.
Parameters:
vDepData=Input, vector holding input dependent data
mIndepData=Input, matrix holding input independent data
trMR: Input settings and Output results, tree having the nodes:
trMR.GUI.Fit.ThroughZero.nVal // Force fit to pass thru zero
trMR.Calculation.Parameters.P1.Name.strVal // Parameter Name...Y-Intercept
trMR.Calculation.Parameters.P1.Value.dVal // Parameter Value
trMR.Calculation.Parameters.P1.Error.dVal // Parameter Error
trMR.Calculation.Parameters.P1.tValue.dVal // Parameter t-Value
trMR.Calculation.Parameters.P1.Prob.dVal // Parameter Prob > |t|
trMR.Calculation.Parameters.P2.Name.strVal // Parameter Name...<Dataset Name>
trMR.Calculation.Parameters.P2.Value.dVal
trMR.Calculation.Parameters.P2.Error.dVal
trMR.Calculation.Parameters.P2.tValue.dVal
trMR.Calculation.Parameters.P2.Prob.dVal
etc. depending on number of dependent variables
trMR.Calculation.Stat.Rvalue.dVal // R-value
trMR.Calculation.Stat.RSqCOD.dVal // RSq-COD value
trMR.Calculation.Stat.AdjRSq.dVal // Adjusted RSq
trMR.Calculation.Stat.RMSESD.dVal // RMSE-Standard Dev.
trMR.Calculation.Stat.N.nVal // No. of points in fit
trMR.Calculation.ANOVA.Model.DOF.nVal // Model DOF
trMR.Calculation.ANOVA.Model.SSq.dVal // Model SS
trMR.Calculation.ANOVA.Model.MeanSq.dVal // Model Mean Sq
trMR.Calculation.ANOVA.Error.DOF.nVal // Error DOF
trMR.Calculation.ANOVA.Error.SSq.dVal // Error SS
trMR.Calculation.ANOVA.Error.MeanSq.dVal // Error Mean Sq
trMR.Calculation.ANOVA.Total.DOF.nVal // Total degrees of freedom
trMR.Calculation.ANOVA.Total.SSq.dVal // Total SS
trMR.Calculation.ANOVA.Fvalue.dVal // F-value
trMR.Calculation.ANOVA.Pvalue.dVal // P-value
vWeightData=Input, Optional vector containing input weighting data if used
mCov=Output, the variance-covariance matrix
Return:
Returns ERROR_NO_ERROR on successful exit and an error code on failure.
NAG error codes: Negative 1 * NAG error code
ERROR_INVALID_TREENODE: Invalid TreeNode
ERROR_TO_FEW_PTS: To few data points
ERROR_UNEQUAL_N: Unequal N where equal N is required
*/
int stat_multiple_linear_regression(const vector& vDepData, const matrix& mIndepData,
TreeNode& trMR, const vector* vWeightData = NULL, matrix* mCov = NULL );
/** >Statistics
Perform a polynomial fit on the input curve.
Parameters:
crvData=Input, the X and Y coordinates of the sample data
mResultCurves=Output, the result curves where
Col(0) is the X coordinates of the fit line
Col(1) is the Y coordinates of the fit line
Col(2) is the lower confidence band
Col(3) is the upper confidence band
Col(4) is the lower prediction band
Col(5) is the upper prediction band
trPR=Input settings & Output results of the linear regression
trPR.GUI.Fit.PolynomialOrder.nVal // The order of the polynomial
trPR.GUI.Fit.ThroughZero.nVal // Force fit to pass thru zero
trPR.GUI.Fit.ErrBarWeight.nVal // Use error column for wt - use (1/err^2) as wt factor
trPR.GUI.Fit.UseReducedChiSq.nVal // Scale parameter errors with reduced chisqr
trPR.GUI.ResultCurves.Points.nVal // Number of points in fit curve
trPR.GUI.ResultCurves.ConfBands.nVal // Create confidence bands - if not set, then matrix will have empty columns
trPR.GUI.ResultCurves.PredBands.nVal // Create prediction bands - if not set, then matrix will have empty columns
trPR.GUI.ResultCurves.Confidence.dVal // Confidence value to be used
trPR.Calculation.Control.UseDataXRange.nVal // Option = 1 to use data range, = 0 to use X1, X2
trPR.Calculation.Control.X1.dVal // Default X minimum for fit curve
trPR.Calculation.Control.X2.dVal // Default X maximum for fit curve
trPR.Calculation.Parameters.P1.Name.strVal // Parameter Name...A
trPR.Calculation.Parameters.P1.Value.dVal // Parameter Value
trPR.Calculation.Parameters.P1.Error.dVal // Parameter Error
trPR.Calculation.Parameters.P1.Vary.nVal // Parameter Vary
trPR.Calculation.Parameters.P1.tValue.dVal // Parameter t-Value
trPR.Calculation.Parameters.P1.LCI.dVal // Parameter LCI
trPR.Calculation.Parameters.P1.UCI.dVal // Parameter UCI
trPR.Calculation.Parameters.P1.Prob.dVal // Parameter Prob > |t|
trPR.Calculation.Parameters.P2.Name.strVal // Parameter Name...B1
trPR.Calculation.Parameters.P2.Value.dVal
trPR.Calculation.Parameters.P2.Error.dVal
trPR.Calculation.Parameters.P2.Vary.nVal
trPR.Calculation.Parameters.P2.tValue.dVal
trPR.Calculation.Parameters.P2.LCI.dVal
trPR.Calculation.Parameters.P2.UCI.dVal
trPR.Calculation.Parameters.P2.Prob.dVal
etc. per order of Polynomial
trPR.Calculation.Stat.Rvalue.dVal // R-value
trPR.Calculation.Stat.RSqCOD.dVal // RSq-COD value
trPR.Calculation.Stat.AdjRSq.dVal // Adjusted RSq
trPR.Calculation.Stat.RMSESD.dVal // RMSE-Standard Dev.
trPR.Calculation.Stat.N.nVal // No. of points in fit
trPR.Calculation.ANOVA.Model.DOF.nVal // Model DOF
trPR.Calculation.ANOVA.Model.SSq.dVal // Model SS
trPR.Calculation.ANOVA.Model.MeanSq.dVal // Model Mean Sq
trPR.Calculation.ANOVA.Error.DOF.nVal // Error DOF
trPR.Calculation.ANOVA.Error.SSq.dVal // Error SS
trPR.Calculation.ANOVA.Error.MeanSq.dVal // Error Mean Sq
trPR.Calculation.ANOVA.Total.DOF.dVal // Total degrees of freedom
trPR.Calculation.ANOVA.Total.SSq.dVal // Total SS
trPR.Calculation.ANOVA.Fvalue.dVal // F-value
trPR.Calculation.ANOVA.Pvalue.dVal // P-value
Return:
Returns ERROR_NO_ERROR on successful exit and an error code on failure.
NAG error codes: Negative 1 * NAG error code
ERROR_INVALID_CURVE: Invalid input curve
ERROR_INVALID_TREENODE: Invalid TreeNode
ERROR_TO_FEW_PTS: Not enough points in the curve
ERROR_SETTINGS: Polynomial order invalid
*/
int stat_polynomial_fit(const curvebase& crvData, matrix& mResultCurves, TreeNode& trPR);
#endif // _STAT_UTILS_H⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -