Logo ROOT  
Reference Guide
 
Loading...
Searching...
No Matches
TMultiDimFit Class Reference

Multidimensional Fits in ROOT.

Overview

A common problem encountered in different fields of applied science is to find an expression for one physical quantity in terms of several others, which are directly measurable.

An example in high energy physics is the evaluation of the momentum of a charged particle from the observation of its trajectory in a magnetic field. The problem is to relate the momentum of the particle to the observations, which may consists of positional measurements at intervals along the particle trajectory.

The exact functional relationship between the measured quantities (e.g., the space-points) and the dependent quantity (e.g., the momentum) is in general not known, but one possible way of solving the problem, is to find an expression which reliably approximates the dependence of the momentum on the observations.

This explicit function of the observations can be obtained by a least squares fitting procedure applied to a representative sample of the data, for which the dependent quantity (e.g., momentum) and the independent observations are known. The function can then be used to compute the quantity of interest for new observations of the independent variables.

This class TMultiDimFit implements such a procedure in ROOT. It is largely based on the CERNLIB MUDIFI package 2. Though the basic concepts are still sound, and therefore kept, a few implementation details have changed, and this class can take advantage of MINUIT 4 to improve the errors of the fitting, thanks to the class TMinuit.

In 5 and 6 H. Wind demonstrates the utility of this procedure in the context of tracking, magnetic field parameterisation, and so on. The outline of the method used in this class is based on Winds discussion, and I refer these two excellents text for more information.

And example of usage is given in multidimfit.C.

The Method

Let \( D \) by the dependent quantity of interest, which depends smoothly on the observable quantities \( x_1, \ldots, x_N \) which we'll denote by \(\mathbf{x}\). Given a training sample of \( M\) tuples of the form, (TMultiDimFit::AddRow)

\[ \left(\mathbf{x}_j, D_j, E_j\right)\quad, \]

where \(\mathbf{x}_j = (x_{1,j},\ldots,x_{N,j})\) are \( N\) independent variables, \( D_j\) is the known, quantity dependent at \(\mathbf{x}_j\) and \( E_j\) is the square error in \( D_j\), the class will try to find the parameterization

\[ D_p(\mathbf{x}) = \sum_{l=1}^{L} c_l \prod_{i=1}^{N} p_{li}\left(x_i\right) = \sum_{l=1}^{L} c_l F_l(\mathbf{x}) \]

such that

\[ S \equiv \sum_{j=1}^{M} \left(D_j - D_p\left(\mathbf{x}_j\right)\right)^2 \]

is minimal. Here \(p_{li}(x_i)\) are monomials, or Chebyshev or Legendre polynomials, labelled \(l = 1, \ldots, L\), in each variable \( x_i\), \( i=1, \ldots, N\).

So what TMultiDimFit does, is to determine the number of terms \( L\), and then \( L\) terms (or functions) \( F_l\), and the \( L\) coefficients \( c_l\), so that \( S\) is minimal (TMultiDimFit::FindParameterization).

Of course it's more than a little unlikely that \( S\) will ever become exact zero as a result of the procedure outlined below. Therefore, the user is asked to provide a minimum relative error \( \epsilon\) (TMultiDimFit::SetMinRelativeError), and \( S\) will be considered minimized when

\[ R = \frac{S}{\sum_{j=1}^M D_j^2} < \epsilon \]

Optionally, the user may impose a functional expression by specifying the powers of each variable in \( L\) specified functions \( F_1, \ldots,F_L\) (TMultiDimFit::SetPowers). In that case, only the coefficients \( c_l\) is calculated by the class.

Limiting the Number of Terms

As always when dealing with fits, there's a real chance of over fitting. As is well-known, it's always possible to fit an \( N-1\) polynomial in \( x\) to \( N\) points \( (x,y)\) with \(\chi^2 = 0\), but the polynomial is not likely to fit new data at all 1. Therefore, the user is asked to provide an upper limit, \( L_{max}\) to the number of terms in \( D_p\) (TMultiDimFit::SetMaxTerms).

However, since there's an infinite number of \( F_l\) to choose from, the user is asked to give the maximum power. \( P_{max,i}\), of each variable \( x_i\) to be considered in the minimization of \( S\) (TMultiDimFit::SetMaxPowers).

One way of obtaining values for the maximum power in variable \( i\), is to perform a regular fit to the dependent quantity \( D\), using a polynomial only in \( x_i\). The maximum power is \( P_{max,i}\) is then the power that does not significantly improve the one-dimensional least-square fit over \( x_i\) to \( D\) 5.

There are still a huge amount of possible choices for \( F_l\); in fact there are \(\prod_{i=1}^{N} (P_{max,i} + 1)\) possible choices. Obviously we need to limit this. To this end, the user is asked to set a power control limit, \( Q\) (TMultiDimFit::SetPowerLimit), and a function \( F_l\) is only accepted if

\[ Q_l = \sum_{i=1}^{N} \frac{P_{li}}{P_{max,i}} < Q \]

where \( P_{li}\) is the leading power of variable \( x_i\) in function \( F_l\) (TMultiDimFit::MakeCandidates). So the number of functions increase with \( Q\) (1, 2 is fine, 5 is way out).

Gram-Schmidt Orthogonalisation

To further reduce the number of functions in the final expression, only those functions that significantly reduce \( S\) is chosen. What ‘significant’ means, is chosen by the user, and will be discussed below (see 2.3).

The functions \( F_l\) are generally not orthogonal, which means one will have to evaluate all possible \( F_l\)'s over all data-points before finding the most significant 1. We can, however, do better then that. By applying the modified Gram-Schmidt orthogonalisation algorithm [5] [3] to the functions \( F_l\), we can evaluate the contribution to the reduction of \( S\) from each function in turn, and we may delay the actual inversion of the curvature-matrix (TMultiDimFit::MakeGramSchmidt).

So we are let to consider an \( M\times L\) matrix \(\mathsf{F}\), an element of which is given by

\[ f_{jl} = F_j\left(x_{1j} , x_{2j}, \ldots, x_{Nj}\right) = F_l(\mathbf{x}_j)\, \quad\mbox{with}~j=1,2,\ldots,M, \]

where \( j\) labels the \( M\) rows in the training sample and \( l\) labels \( L\) functions of \( N\) variables, and \( L \leq M\). That is, \( f_{jl}\) is the term (or function) numbered \( l\) evaluated at the data point \( j\). We have to normalise \(\mathbf{x}_j\) to \( [-1,1]\) for this to succeed [5] (TMultiDimFit::MakeNormalized). We then define a matrix \(\mathsf{W}\) of which the columns \(\mathbf{w}_j\) are given by

\begin{eqnarray*} \mathbf{w}_1 &=& \mathbf{f}_1 = F_1\left(\mathbf x_1\right)\\ \mathbf{w}_l &=& \mathbf{f}_l - \sum^{l-1}_{k=1} \frac{\mathbf{f}_l \bullet \mathbf{w}_k}{\mathbf{w}_k^2}\mathbf{w}_k\,. \end{eqnarray*}

and \(\mathbf{w}_{l}\) is the component of \(\mathbf{f}_{l} \) orthogonal to \(\mathbf{w}_{1}, \ldots, \mathbf{w}_{l-1}\). Hence we obtain [3],

\[ \mathbf{w}_k\bullet\mathbf{w}_l = 0\quad\mbox{if}~k \neq l\quad. \]

We now take as a new model \(\mathsf{W}\mathbf{a}\). We thus want to minimize

\[ S\equiv \left(\mathbf{D} - \mathsf{W}\mathbf{a}\right)^2\quad, \]

where \(\mathbf{D} = \left(D_1,\ldots,D_M\right)\) is a vector of the dependent quantity in the sample. Differentiation with respect to \( a_j\) gives, using 6,

\[ \mathbf{D}\bullet\mathbf{w}_l - a_l\mathbf{w}_l^2 = 0 \]

or

\[ a_l = \frac{\mathbf{D}_l\bullet\mathbf{w}_l}{\mathbf{w}_l^2} \]

Let \( S_j\) be the sum of squares of residuals when taking \( j\) functions into account. Then

\[ S_l = \left[\mathbf{D} - \sum^l_{k=1} a_k\mathbf{w}_k\right]^2 = \mathbf{D}^2 - 2\mathbf{D} \sum^l_{k=1} a_k\mathbf{w}_k + \sum^l_{k=1} a_k^2\mathbf{w}_k^2 \]

Using 9, we see that

\begin{eqnarray*} S_l &=& \mathbf{D}^2 - 2 \sum^l_{k=1} a_k^2\mathbf{w}_k^2 + \sum^j_{k=1} a_k^2\mathbf{w}_k^2\nonumber\\ &=& \mathbf{D}^2 - \sum^l_{k=1} a_k^2\mathbf{w}_k^2\nonumber\\ &=& \mathbf{D}^2 - \sum^l_{k=1} \frac{\left(\mathbf D\bullet \mathbf w_k\right)}{\mathbf w_k^2} \end{eqnarray*}

So for each new function \( F_l\) included in the model, we get a reduction of the sum of squares of residuals of \(a_l^2\mathbf{w}_l^2\), where \(\mathbf{w}_l\) is given by 4 and \( a_l\) by 9. Thus, using the Gram-Schmidt orthogonalisation, we can decide if we want to include this function in the final model, before* the matrix inversion.

Function Selection Based on Residual

Supposing that \( L-1\) steps of the procedure have been performed, the problem now is to consider the \(L^{\mbox{th}}\) function.

The sum of squares of residuals can be written as

\[ S_L = \textbf{D}^T\bullet\textbf{D} - \sum^L_{l=1}a^2_l\left(\textbf{w}_l^T\bullet\textbf{w}_l\right) \]

where the relation 9 have been taken into account. The contribution of the \(L^{\mbox{th}}\) function to the reduction of S, is given by

\[ \Delta S_L = a^2_L\left(\textbf{w}_L^T\bullet\textbf{w}_L\right) \]

Two test are now applied to decide whether this \(L^{\mbox{th}}\) function is to be included in the final expression, or not.

Test 1

Denoting by \( H_{L-1}\) the subspace spanned by \(\textbf{w}_1,\ldots,\textbf{w}_{L-1}\) the function \(\textbf{w}_L\) is by construction (see 4) the projection of the function \( F_L\) onto the direction perpendicular to \( H_{L-1}\). Now, if the length of \(\textbf{w}_L\) (given by \(\textbf{w}_L\bullet\textbf{w}_L\)) is very small compared to the length of \(\textbf{f}_L\) this new function can not contribute much to the reduction of the sum of squares of residuals. The test consists then in calculating the angle \( \theta \) between the two vectors \(\textbf{w}_L\) \( \textbf {f}_L\) (see also figure 1) and requiring that it's greater* then a threshold value which the user must set (TMultiDimFit::SetMinAngle).

Figure 1: (a) angle \f$\theta\f$ between \f$\textbf{w}_l\f$ and \f$\textbf{f}_L\f$, (b) angle \f$ \phi \f$ between \f$\textbf{w}_L\f$ and \f$\textbf{D}\f$

Test 2

Let \(\textbf{D}\) be the data vector to be fitted. As illustrated in figure 1, the \(L^{\mbox{th}}\) function \(\textbf{w}_L\) will contribute significantly to the reduction of \( S\), if the angle \(\phi^\prime\) between \(\textbf{w}_L\) and \(\textbf{D}\) is smaller than an upper limit \( \phi \), defined by the user (MultiDimFit::SetMaxAngle)

However, the method automatically readjusts the value of this angle while fitting is in progress, in order to make the selection criteria less and less difficult to be fulfilled. The result is that the functions contributing most to the reduction of \( S\) are chosen first (TMultiDimFit::TestFunction).

In case \( \phi \) isn't defined, an alternative method of performing this second test is used: The \(L^{\mbox{th}}\) function \(\textbf{f}_L\) is accepted if (refer also to equation (13))

\[ \Delta S_L > \frac{S_{L-1}}{L_{max}-L} \]

where \( S_{L-1}\) is the sum of the \( L-1\) first residuals from the \( L-1\) functions previously accepted; and \( L_{max}\) is the total number of functions allowed in the final expression of the fit (defined by user).

From this we see, that by restricting \( L_{max}\) – the number of terms in the final model – the fit is more difficult to perform, since the above selection criteria is more limiting.

The more coefficients we evaluate, the more the sum of squares of residuals \( S\) will be reduced. We can evaluate \( S\) before inverting \(\mathsf{B}\) as shown below.

Coefficients and Coefficient Errors

Having found a parameterization, that is the \( F_l\)'s and \( L\), that minimizes \( S\), we still need to determine the coefficients \( c_l\). However, it's a feature of how we choose the significant functions, that the evaluation of the \( c_l\)'s becomes trivial 5. To derive \(\mathbf{c}\), we first note that equation (4) can be written as

\[ \mathsf{F} = \mathsf{W}\mathsf{B} \]

where

\begin{eqnarray*} b_{ij} = \frac{\mathbf{f}_j \bullet \mathbf{w}_i}{\mathbf{w}_i^2} & \mbox{if} & i < j\\ 1 & \mbox{if} & i = j\\ 0 & \mbox{if} & i > j \end{eqnarray*}

Consequently, \(\mathsf{B}\) is an upper triangle matrix, which can be readily inverted. So we now evaluate

\[ \mathsf{F}\mathsf{B}^{-1} = \mathsf{W} \]

The model \(\mathsf{W}\mathbf{a}\) can therefore be written as \((\mathsf{F}\mathsf{B}^{-1})\mathbf{a} = \mathsf{F}(\mathsf{B}^{-1}\mathbf{a})\,.\)

The original model \(\mathsf{F}\mathbf{c}\) is therefore identical with this if

\[ \mathbf{c} = \left(\mathsf{B}^{-1}\mathbf{a}\right) = \left[\mathbf{a}^T\left(\mathsf{B}^{-1}\right)^T\right]^T\,. \]

The reason we use \(\left(\mathsf{B}^{-1}\right)^T\) rather then \(\mathsf{B}^{-1}\) is to save storage, since \(\left(\mathsf{B}^{-1}\right)^T\) can be stored in the same matrix as \(\mathsf{B}\) (TMultiDimFit::MakeCoefficients). The errors in the coefficients is calculated by inverting the curvature matrix of the non-orthogonal functions \( f_{lj}\) [1] (TMultiDimFit::MakeCoefficientErrors).

Considerations

It's important to realize that the training sample should be representative of the problem at hand, in particular along the borders of the region of interest. This is because the algorithm presented here, is a interpolation, rather then a extrapolation 5.

Also, the independent variables \( x_{i}\) need to be linear independent, since the procedure will perform poorly if they are not 5. One can find an linear transformation from ones original variables \( \xi_{i}\) to a set of linear independent variables \( x_{i}\), using a Principal Components Analysis (see TPrincipal), and then use the transformed variable as input to this class [5] 6.

H. Wind also outlines a method for parameterising a multidimensional dependence over a multidimensional set of variables. An example of the method from 5, is a follows (please refer to 5 for a full discussion):

  1. Define \(\mathbf{P} = (P_1, \ldots, P_5)\) are the 5 dependent quantities that define a track.
  2. Compute, for \( M\) different values of \(\mathbf{P}\), the tracks through the magnetic field, and determine the corresponding \(\mathbf{x} = (x_1, \ldots, x_N)\).
  3. Use the simulated observations to determine, with a simple approximation, the values of \(\mathbf{P}_j\). We call these values \(\mathbf{P}^\prime_j, j = 1, \ldots, M\).
  4. Determine from \(\mathbf{x}\) a set of at least five relevant coordinates \(\mathbf{x}^\prime\), using contrains, or alternative:
  5. Perform a Principal Component Analysis (using TPrincipal), and use to get a linear transformation \(\mathbf{x} \rightarrow \mathbf{x}^\prime\), so that \(\mathbf{x}^\prime\) are constrained and linear independent.
  6. Perform a Principal Component Analysis on \(Q_i = P_i / P^\prime_i\, i = 1, \ldots, 5\), to get linear indenpendent (among themselves, but not independent of \(\mathbf{x}\)) quantities \(\mathbf{Q}^\prime\)
  7. For each component \(Q^\prime_i\) make a multidimensional fit, using \(\mathbf{x}^\prime\) as the variables, thus determining a set of coefficients \(\mathbf{c}_i\).

To process data, using this parameterisation, do

  1. Test wether the observation \(\mathbf{x}\) within the domain of the parameterization, using the result from the Principal Component Analysis.
  2. Determine \(\mathbf{P}^\prime\) as before.
  3. Determine \(\mathbf{x}^\prime\) as before.
  4. Use the result of the fit to determine \(\mathbf{Q}^\prime\).
  5. Transform back to \(\mathbf{P}\) from \(\mathbf{Q}^\prime\), using the result from the Principal Component Analysis.

Testing the parameterization

The class also provides functionality for testing the, over the training sample, found parameterization (TMultiDimFit::Fit). This is done by passing the class a test sample of \( M_t\) tuples of the form \((\mathbf{x}_{t,j},D_{t,j}, E_{t,j})\), where \(\mathbf{x}_{t,j}\) are the independent variables, \( D_{t,j}\) the known, dependent quantity, and \( E_{t,j}\) is the square error in \( D_{t,j}\) (TMultiDimFit::AddTestRow).

The parameterization is then evaluated at every \(\mathbf{x}_t\) in the test sample, and

\[ S_t \equiv \sum_{j=1}^{M_t} \left(D_{t,j} - D_p\left(\mathbf{x}_{t,j}\right)\right)^2 \]

is evaluated. The relative error over the test sample

\[ R_t = \frac{S_t}{\sum_{j=1}^{M_t} D_{t,j}^2} \]

should not be to low or high compared to \( R\) from the training sample. Also, multiple correlation coefficient from both samples should be fairly close, otherwise one of the samples is not representative of the problem. A large difference in the reduced \( \chi^2\) over the two samples indicate an over fit, and the maximum number of terms in the parameterisation should be reduced.

It's possible to use 4 to further improve the fit, using the test sample.

Christian Holm

Bibliography

  • Philip R. Bevington and D. Keith Robinson. Data Reduction and Error Analysis for the Physical Sciences. McGraw-Hill, 2 edition, 1992.
  • R. Brun et al. Long writeup DD/75-23, CERN, 1980.
  • Gene H. Golub and Charles F. van Loan. Matrix Computations. John Hopkins University Press, Baltimore, 3 edition, 1996.
  • F. James. Minuit. Long writeup D506, CERN, 1998.
  • H. Wind. Function parameterization. Proceedings of the 1972 CERN Computing and Data Processing School, volume 72-21 of Yellow report. CERN, 1972.
  • H. Wind. 1. principal component analysis, 2. pattern recognition for track finding, 3. interpolation and functional representation. Yellow report EP/81-12, CERN, 1981.

Definition at line 15 of file TMultiDimFit.h.

Public Types

enum  EMDFPolyType { kMonomials , kChebyshev , kLegendre }
 
- Public Types inherited from TObject
enum  {
  kIsOnHeap = 0x01000000 , kNotDeleted = 0x02000000 , kZombie = 0x04000000 , kInconsistent = 0x08000000 ,
  kBitMask = 0x00ffffff
}
 
enum  { kSingleKey = (1ULL << ( 0 )) , kOverwrite = (1ULL << ( 1 )) , kWriteDelete = (1ULL << ( 2 )) }
 
enum  EDeprecatedStatusBits { kObjInCanvas = (1ULL << ( 3 )) }
 
enum  EStatusBits {
  kCanDelete = (1ULL << ( 0 )) , kMustCleanup = (1ULL << ( 3 )) , kIsReferenced = (1ULL << ( 4 )) , kHasUUID = (1ULL << ( 5 )) ,
  kCannotPick = (1ULL << ( 6 )) , kNoContextMenu = (1ULL << ( 8 )) , kInvalidObject = (1ULL << ( 13 ))
}
 

Public Member Functions

 TMultiDimFit ()
 Empty CTOR. Do not use.
 
 TMultiDimFit (Int_t dimension, EMDFPolyType type=kMonomials, Option_t *option="")
 Constructor Second argument is the type of polynomials to use in parameterisation, one of: TMultiDimFit::kMonomials TMultiDimFit::kChebyshev TMultiDimFit::kLegendre.
 
 ~TMultiDimFit () override
 Destructor.
 
virtual void AddRow (const Double_t *x, Double_t D, Double_t E=0)
 Add a row consisting of fNVariables independent variables, the known, dependent quantity, and optionally, the square error in the dependent quantity, to the training sample to be used for the parameterization.
 
virtual void AddTestRow (const Double_t *x, Double_t D, Double_t E=0)
 Add a row consisting of fNVariables independent variables, the known, dependent quantity, and optionally, the square error in the dependent quantity, to the test sample to be used for the test of the parameterization.
 
void Browse (TBrowser *b) override
 Browse the TMultiDimFit object in the TBrowser.
 
void Clear (Option_t *option="") override
 Clear internal structures and variables.
 
void Draw (Option_t *="d") override
 Default Draw method for all objects.
 
virtual Double_t Eval (const Double_t *x, const Double_t *coeff=nullptr) const
 Evaluate parameterization at point x.
 
virtual Double_t EvalError (const Double_t *x, const Double_t *coeff=nullptr) const
 Evaluate parameterization error at point x.
 
virtual void FindParameterization (Option_t *option="")
 Find the parameterization.
 
virtual void Fit (Option_t *option="")
 Try to fit the found parameterisation to the test sample.
 
Double_t GetChi2 () const
 
const TVectorDGetCoefficients () const
 
const TVectorDGetCoefficientsRMS () const
 
const TMatrixDGetCorrelationMatrix () const
 
Double_t GetError () const
 
Int_tGetFunctionCodes () const
 
const TMatrixDGetFunctions () const
 
virtual TListGetHistograms () const
 
Double_t GetMaxAngle () const
 
Int_t GetMaxFunctions () const
 
Int_tGetMaxPowers () const
 
Double_t GetMaxQuantity () const
 
Int_t GetMaxStudy () const
 
Int_t GetMaxTerms () const
 
const TVectorDGetMaxVariables () const
 
Double_t GetMeanQuantity () const
 
const TVectorDGetMeanVariables () const
 
Double_t GetMinAngle () const
 
Double_t GetMinQuantity () const
 
Double_t GetMinRelativeError () const
 
const TVectorDGetMinVariables () const
 
Int_t GetNCoefficients () const
 
Int_t GetNVariables () const
 
Int_t GetPolyType () const
 
Int_tGetPowerIndex () const
 
Double_t GetPowerLimit () const
 
const Int_tGetPowers () const
 
Double_t GetPrecision () const
 
const TVectorDGetQuantity () const
 
Double_t GetResidualMax () const
 
Int_t GetResidualMaxRow () const
 
Double_t GetResidualMin () const
 
Int_t GetResidualMinRow () const
 
Double_t GetResidualSumSq () const
 
Double_t GetRMS () const
 
Int_t GetSampleSize () const
 
const TVectorDGetSqError () const
 
Double_t GetSumSqAvgQuantity () const
 
Double_t GetSumSqQuantity () const
 
Double_t GetTestError () const
 
Double_t GetTestPrecision () const
 
const TVectorDGetTestQuantity () const
 
Int_t GetTestSampleSize () const
 
const TVectorDGetTestSqError () const
 
const TVectorDGetTestVariables () const
 
const TVectorDGetVariables () const
 
TClassIsA () const override
 
Bool_t IsFolder () const override
 Returns kTRUE in case object contains browsable objects (like containers or lists of other objects).
 
virtual Double_t MakeChi2 (const Double_t *coeff=nullptr)
 Calculate Chi square over either the test sample.
 
virtual void MakeCode (const char *functionName="MDF", Option_t *option="")
 Generate the file <filename> with .C appended if argument doesn't end in .cxx or .C.
 
virtual void MakeHistograms (Option_t *option="A")
 Make histograms of the result of the analysis.
 
virtual void MakeMethod (const Char_t *className="MDF", Option_t *option="")
 Generate the file <classname>MDF.cxx which contains the implementation of the method:
 
void Print (Option_t *option="ps") const override
 Print statistics etc.
 
void SetBinVarX (Int_t nbbinvarx)
 
void SetBinVarY (Int_t nbbinvary)
 
void SetMaxAngle (Double_t angle=0)
 Set the max angle (in degrees) between the initial data vector to be fitted, and the new candidate function to be included in the fit.
 
void SetMaxFunctions (Int_t n)
 
void SetMaxPowers (const Int_t *powers)
 Set the maximum power to be considered in the fit for each variable.
 
void SetMaxStudy (Int_t n)
 
void SetMaxTerms (Int_t terms)
 
void SetMinAngle (Double_t angle=1)
 Set the min angle (in degrees) between a new candidate function and the subspace spanned by the previously accepted functions.
 
void SetMinRelativeError (Double_t error)
 Set the acceptable relative error for when sum of square residuals is considered minimized.
 
void SetPowerLimit (Double_t limit=1e-3)
 Set the user parameter for the function selection.
 
virtual void SetPowers (const Int_t *powers, Int_t terms)
 Define a user function.
 
void Streamer (TBuffer &) override
 Stream an object of class TObject.
 
void StreamerNVirtual (TBuffer &ClassDef_StreamerNVirtual_b)
 
- Public Member Functions inherited from TNamed
 TNamed ()
 
 TNamed (const char *name, const char *title)
 
 TNamed (const TNamed &named)
 TNamed copy ctor.
 
 TNamed (const TString &name, const TString &title)
 
virtual ~TNamed ()
 TNamed destructor.
 
void Clear (Option_t *option="") override
 Set name and title to empty strings ("").
 
TObjectClone (const char *newname="") const override
 Make a clone of an object using the Streamer facility.
 
Int_t Compare (const TObject *obj) const override
 Compare two TNamed objects.
 
void Copy (TObject &named) const override
 Copy this to obj.
 
virtual void FillBuffer (char *&buffer)
 Encode TNamed into output buffer.
 
const char * GetName () const override
 Returns name of object.
 
const char * GetTitle () const override
 Returns title of object.
 
ULong_t Hash () const override
 Return hash value for this object.
 
TClassIsA () const override
 
Bool_t IsSortable () const override
 
void ls (Option_t *option="") const override
 List TNamed name and title.
 
TNamedoperator= (const TNamed &rhs)
 TNamed assignment operator.
 
void Print (Option_t *option="") const override
 Print TNamed name and title.
 
virtual void SetName (const char *name)
 Set the name of the TNamed.
 
virtual void SetNameTitle (const char *name, const char *title)
 Set all the TNamed parameters (name and title).
 
virtual void SetTitle (const char *title="")
 Set the title of the TNamed.
 
virtual Int_t Sizeof () const
 Return size of the TNamed part of the TObject.
 
void Streamer (TBuffer &) override
 Stream an object of class TObject.
 
void StreamerNVirtual (TBuffer &ClassDef_StreamerNVirtual_b)
 
- Public Member Functions inherited from TObject
 TObject ()
 TObject constructor.
 
 TObject (const TObject &object)
 TObject copy ctor.
 
virtual ~TObject ()
 TObject destructor.
 
void AbstractMethod (const char *method) const
 Use this method to implement an "abstract" method that you don't want to leave purely abstract.
 
virtual void AppendPad (Option_t *option="")
 Append graphics object to current pad.
 
ULong_t CheckedHash ()
 Check and record whether this class has a consistent Hash/RecursiveRemove setup (*) and then return the regular Hash value for this object.
 
virtual const char * ClassName () const
 Returns name of class to which the object belongs.
 
virtual void Delete (Option_t *option="")
 Delete this object.
 
virtual Int_t DistancetoPrimitive (Int_t px, Int_t py)
 Computes distance from point (px,py) to the object.
 
virtual void DrawClass () const
 Draw class inheritance tree of the class to which this object belongs.
 
virtual TObjectDrawClone (Option_t *option="") const
 Draw a clone of this object in the current selected pad with: gROOT->SetSelectedPad(c1).
 
virtual void Dump () const
 Dump contents of object on stdout.
 
virtual void Error (const char *method, const char *msgfmt,...) const
 Issue error message.
 
virtual void Execute (const char *method, const char *params, Int_t *error=nullptr)
 Execute method on this object with the given parameter string, e.g.
 
virtual void Execute (TMethod *method, TObjArray *params, Int_t *error=nullptr)
 Execute method on this object with parameters stored in the TObjArray.
 
virtual void ExecuteEvent (Int_t event, Int_t px, Int_t py)
 Execute action corresponding to an event at (px,py).
 
virtual void Fatal (const char *method, const char *msgfmt,...) const
 Issue fatal error message.
 
virtual TObjectFindObject (const char *name) const
 Must be redefined in derived classes.
 
virtual TObjectFindObject (const TObject *obj) const
 Must be redefined in derived classes.
 
virtual Option_tGetDrawOption () const
 Get option used by the graphics system to draw this object.
 
virtual const char * GetIconName () const
 Returns mime type name of object.
 
virtual char * GetObjectInfo (Int_t px, Int_t py) const
 Returns string containing info about the object at position (px,py).
 
virtual Option_tGetOption () const
 
virtual UInt_t GetUniqueID () const
 Return the unique object id.
 
virtual Bool_t HandleTimer (TTimer *timer)
 Execute action in response of a timer timing out.
 
Bool_t HasInconsistentHash () const
 Return true is the type of this object is known to have an inconsistent setup for Hash and RecursiveRemove (i.e.
 
virtual void Info (const char *method, const char *msgfmt,...) const
 Issue info message.
 
virtual Bool_t InheritsFrom (const char *classname) const
 Returns kTRUE if object inherits from class "classname".
 
virtual Bool_t InheritsFrom (const TClass *cl) const
 Returns kTRUE if object inherits from TClass cl.
 
virtual void Inspect () const
 Dump contents of this object in a graphics canvas.
 
void InvertBit (UInt_t f)
 
Bool_t IsDestructed () const
 IsDestructed.
 
virtual Bool_t IsEqual (const TObject *obj) const
 Default equal comparison (objects are equal if they have the same address in memory).
 
R__ALWAYS_INLINE Bool_t IsOnHeap () const
 
R__ALWAYS_INLINE Bool_t IsZombie () const
 
void MayNotUse (const char *method) const
 Use this method to signal that a method (defined in a base class) may not be called in a derived class (in principle against good design since a child class should not provide less functionality than its parent, however, sometimes it is necessary).
 
virtual Bool_t Notify ()
 This method must be overridden to handle object notification (the base implementation is no-op).
 
void Obsolete (const char *method, const char *asOfVers, const char *removedFromVers) const
 Use this method to declare a method obsolete.
 
void operator delete (void *ptr)
 Operator delete.
 
void operator delete[] (void *ptr)
 Operator delete [].
 
void * operator new (size_t sz)
 
void * operator new (size_t sz, void *vp)
 
void * operator new[] (size_t sz)
 
void * operator new[] (size_t sz, void *vp)
 
TObjectoperator= (const TObject &rhs)
 TObject assignment operator.
 
virtual void Paint (Option_t *option="")
 This method must be overridden if a class wants to paint itself.
 
virtual void Pop ()
 Pop on object drawn in a pad to the top of the display list.
 
virtual Int_t Read (const char *name)
 Read contents of object with specified name from the current directory.
 
virtual void RecursiveRemove (TObject *obj)
 Recursively remove this object from a list.
 
void ResetBit (UInt_t f)
 
virtual void SaveAs (const char *filename="", Option_t *option="") const
 Save this object in the file specified by filename.
 
virtual void SavePrimitive (std::ostream &out, Option_t *option="")
 Save a primitive as a C++ statement(s) on output stream "out".
 
void SetBit (UInt_t f)
 
void SetBit (UInt_t f, Bool_t set)
 Set or unset the user status bits as specified in f.
 
virtual void SetDrawOption (Option_t *option="")
 Set drawing option for object.
 
virtual void SetUniqueID (UInt_t uid)
 Set the unique object id.
 
void StreamerNVirtual (TBuffer &ClassDef_StreamerNVirtual_b)
 
virtual void SysError (const char *method, const char *msgfmt,...) const
 Issue system error message.
 
R__ALWAYS_INLINE Bool_t TestBit (UInt_t f) const
 
Int_t TestBits (UInt_t f) const
 
virtual void UseCurrentStyle ()
 Set current style settings in this object This function is called when either TCanvas::UseCurrentStyle or TROOT::ForceStyle have been invoked.
 
virtual void Warning (const char *method, const char *msgfmt,...) const
 Issue warning message.
 
virtual Int_t Write (const char *name=nullptr, Int_t option=0, Int_t bufsize=0)
 Write this object to the current directory.
 
virtual Int_t Write (const char *name=nullptr, Int_t option=0, Int_t bufsize=0) const
 Write this object to the current directory.
 

Static Public Member Functions

static TClassClass ()
 
static const char * Class_Name ()
 
static constexpr Version_t Class_Version ()
 
static const char * DeclFileName ()
 
static TMultiDimFitInstance ()
 Return the static instance.
 
- Static Public Member Functions inherited from TNamed
static TClassClass ()
 
static const char * Class_Name ()
 
static constexpr Version_t Class_Version ()
 
static const char * DeclFileName ()
 
- Static Public Member Functions inherited from TObject
static TClassClass ()
 
static const char * Class_Name ()
 
static constexpr Version_t Class_Version ()
 
static const char * DeclFileName ()
 
static Longptr_t GetDtorOnly ()
 Return destructor only flag.
 
static Bool_t GetObjectStat ()
 Get status of object stat flag.
 
static void SetDtorOnly (void *obj)
 Set destructor only flag.
 
static void SetObjectStat (Bool_t stat)
 Turn on/off tracking of objects in the TObjectTable.
 

Protected Member Functions

virtual Double_t EvalControl (const Int_t *powers) const
 PRIVATE METHOD: Calculate the control parameter from the passed powers.
 
virtual Double_t EvalFactor (Int_t p, Double_t x) const
 PRIVATE METHOD: Evaluate function with power p at variable value x.
 
virtual void MakeCandidates ()
 PRIVATE METHOD: Create list of candidate functions for the parameterisation.
 
virtual void MakeCoefficientErrors ()
 PRIVATE METHOD: Compute the errors on the coefficients.
 
virtual void MakeCoefficients ()
 PRIVATE METHOD: Invert the model matrix B, and compute final coefficients.
 
virtual void MakeCorrelation ()
 PRIVATE METHOD: Compute the correlation matrix.
 
virtual Double_t MakeGramSchmidt (Int_t function)
 PRIVATE METHOD: Make Gram-Schmidt orthogonalisation.
 
virtual void MakeNormalized ()
 PRIVATE METHOD: Normalize data to the interval [-1;1].
 
virtual void MakeParameterization ()
 PRIVATE METHOD: Find the parameterization over the training sample.
 
virtual void MakeRealCode (const char *filename, const char *classname, Option_t *option="")
 PRIVATE METHOD: This is the method that actually generates the code for the evaluation the parameterization on some point.
 
virtual Bool_t Select (const Int_t *iv)
 Selection method.
 
virtual Bool_t TestFunction (Double_t squareResidual, Double_t dResidur)
 PRIVATE METHOD: Test whether the currently considered function contributes to the fit.
 
- Protected Member Functions inherited from TObject
virtual void DoError (int level, const char *location, const char *fmt, va_list va) const
 Interface to ErrorHandler (protected).
 
void MakeZombie ()
 

Protected Attributes

Int_t fBinVarX
 Number of bin in independent variables.
 
Int_t fBinVarY
 Number of bin in dependent variables.
 
Double_t fChi2
 Chi square of fit.
 
TVectorD fCoefficients
 Vector of the final coefficients.
 
TVectorD fCoefficientsRMS
 Vector of RMS of coefficients.
 
Double_t fCorrelationCoeff
 Multi Correlation coefficient.
 
TMatrixD fCorrelationMatrix
 Correlation matrix.
 
Double_t fError
 Error from parametrization.
 
TVirtualFitterfFitter
 
Int_tfFunctionCodes
 [fMaxFunctions] acceptance code
 
TMatrixD fFunctions
 Functions evaluated over sample.
 
Byte_t fHistogramMask
 Bit pattern of histograms used.
 
TListfHistograms
 List of histograms.
 
Bool_t fIsUserFunction
 Flag for user defined function.
 
Bool_t fIsVerbose
 
Double_t fMaxAngle
 Max angle for accepting new function.
 
Int_t fMaxFuncNV
 fMaxFunctions*fNVariables
 
Int_t fMaxFunctions
 max number of functions
 
Int_tfMaxPowers
 [fNVariables] maximum powers
 
Int_tfMaxPowersFinal
 [fNVariables] maximum powers from fit;
 
Double_t fMaxQuantity
 Max value of dependent quantity.
 
Double_t fMaxResidual
 Max residual value.
 
Int_t fMaxResidualRow
 Row giving max residual.
 
Int_t fMaxStudy
 max functions to study
 
Int_t fMaxTerms
 Max terms expected in final expr.
 
TVectorD fMaxVariables
 max value of independent variables
 
Double_t fMeanQuantity
 Mean of dependent quantity.
 
TVectorD fMeanVariables
 mean value of independent variables
 
Double_t fMinAngle
 Min angle for accepting new function.
 
Double_t fMinQuantity
 Min value of dependent quantity.
 
Double_t fMinRelativeError
 Min relative error accepted.
 
Double_t fMinResidual
 Min residual value.
 
Int_t fMinResidualRow
 Row giving min residual.
 
TVectorD fMinVariables
 min value of independent variables
 
Int_t fNCoefficients
 Dimension of model coefficients.
 
Int_t fNVariables
 Number of independent variables.
 
TVectorD fOrthCoefficients
 The model coefficients.
 
TMatrixD fOrthCurvatureMatrix
 Model matrix.
 
TVectorD fOrthFunctionNorms
 Norm of the evaluated functions.
 
TMatrixD fOrthFunctions
 As above, but orthogonalised.
 
Int_t fParameterisationCode
 Exit code of parameterisation.
 
EMDFPolyType fPolyType
 Fit object (MINUIT)
 
Int_tfPowerIndex
 [fMaxTerms] Index of accepted powers
 
Double_t fPowerLimit
 Control parameter.
 
Int_tfPowers
 [fMaxFuncNV] where fMaxFuncNV = fMaxFunctions*fNVariables
 
Double_t fPrecision
 Relative precision of param.
 
TVectorD fQuantity
 Training sample, dependent quantity.
 
TVectorD fResiduals
 Vector of the final residuals.
 
Double_t fRMS
 Root mean square of fit.
 
Int_t fSampleSize
 Size of training sample.
 
Bool_t fShowCorrelation
 print correlation matrix
 
TVectorD fSqError
 Training sample, error in quantity.
 
Double_t fSumSqAvgQuantity
 Sum of squares away from mean.
 
Double_t fSumSqQuantity
 SumSquare of dependent quantity.
 
Double_t fSumSqResidual
 Sum of Square residuals.
 
Double_t fTestCorrelationCoeff
 Multi Correlation coefficient.
 
Double_t fTestError
 Error from test.
 
Double_t fTestPrecision
 Relative precision of test.
 
TVectorD fTestQuantity
 Test sample, dependent quantity.
 
Int_t fTestSampleSize
 Size of test sample.
 
TVectorD fTestSqError
 Test sample, Error in quantity.
 
TVectorD fTestVariables
 Test sample, independent variables.
 
TVectorD fVariables
 Training sample, independent variables.
 
- Protected Attributes inherited from TNamed
TString fName
 
TString fTitle
 

Static Private Attributes

static TMultiDimFitfgInstance = nullptr
 

Additional Inherited Members

- Protected Types inherited from TObject
enum  { kOnlyPrepStep = (1ULL << ( 3 )) }
 

#include <TMultiDimFit.h>

Inheritance diagram for TMultiDimFit:
[legend]

Member Enumeration Documentation

◆ EMDFPolyType

Enumerator
kMonomials 
kChebyshev 
kLegendre 

Definition at line 18 of file TMultiDimFit.h.

Constructor & Destructor Documentation

◆ TMultiDimFit() [1/2]

TMultiDimFit::TMultiDimFit ( )

Empty CTOR. Do not use.

Definition at line 433 of file TMultiDimFit.cxx.

◆ TMultiDimFit() [2/2]

TMultiDimFit::TMultiDimFit ( Int_t  dimension,
EMDFPolyType  type = kMonomials,
Option_t option = "" 
)

Constructor Second argument is the type of polynomials to use in parameterisation, one of: TMultiDimFit::kMonomials TMultiDimFit::kChebyshev TMultiDimFit::kLegendre.

Options: K Compute (k)correlation matrix V Be verbose

Default is no options.

Definition at line 508 of file TMultiDimFit.cxx.

◆ ~TMultiDimFit()

TMultiDimFit::~TMultiDimFit ( )
override

Destructor.

Definition at line 586 of file TMultiDimFit.cxx.

Member Function Documentation

◆ AddRow()

void TMultiDimFit::AddRow ( const Double_t x,
Double_t  D,
Double_t  E = 0 
)
virtual

Add a row consisting of fNVariables independent variables, the known, dependent quantity, and optionally, the square error in the dependent quantity, to the training sample to be used for the parameterization.

The mean of the variables and quantity is calculated on the fly, as outlined in TPrincipal::AddRow. This sample should be representative of the problem at hand. Please note, that if no error is given Poisson statistics is assumed and the square error is set to the value of dependent quantity. See also the class description

Definition at line 611 of file TMultiDimFit.cxx.

◆ AddTestRow()

void TMultiDimFit::AddTestRow ( const Double_t x,
Double_t  D,
Double_t  E = 0 
)
virtual

Add a row consisting of fNVariables independent variables, the known, dependent quantity, and optionally, the square error in the dependent quantity, to the test sample to be used for the test of the parameterization.

This sample needn't be representative of the problem at hand. Please note, that if no error is given Poisson statistics is assumed and the square error is set to the value of dependent quantity. See also the class description

Definition at line 690 of file TMultiDimFit.cxx.

◆ Browse()

void TMultiDimFit::Browse ( TBrowser b)
overridevirtual

Browse the TMultiDimFit object in the TBrowser.

Reimplemented from TObject.

Definition at line 737 of file TMultiDimFit.cxx.

◆ Class()

static TClass * TMultiDimFit::Class ( )
static
Returns
TClass describing this class

◆ Class_Name()

static const char * TMultiDimFit::Class_Name ( )
static
Returns
Name of this class

◆ Class_Version()

static constexpr Version_t TMultiDimFit::Class_Version ( )
inlinestaticconstexpr
Returns
Version of this class

Definition at line 207 of file TMultiDimFit.h.

◆ Clear()

void TMultiDimFit::Clear ( Option_t option = "")
overridevirtual

Clear internal structures and variables.

Reimplemented from TObject.

Definition at line 789 of file TMultiDimFit.cxx.

◆ DeclFileName()

static const char * TMultiDimFit::DeclFileName ( )
inlinestatic
Returns
Name of the file containing the class declaration

Definition at line 207 of file TMultiDimFit.h.

◆ Draw()

void TMultiDimFit::Draw ( Option_t option = "d")
inlineoverridevirtual

Default Draw method for all objects.

Reimplemented from TObject.

Definition at line 134 of file TMultiDimFit.h.

◆ Eval()

Double_t TMultiDimFit::Eval ( const Double_t x,
const Double_t coeff = nullptr 
) const
virtual

Evaluate parameterization at point x.

Optional argument coeff is a vector of coefficients for the parameterisation, fNCoefficients elements long.

Definition at line 876 of file TMultiDimFit.cxx.

◆ EvalControl()

Double_t TMultiDimFit::EvalControl ( const Int_t powers) const
protectedvirtual

PRIVATE METHOD: Calculate the control parameter from the passed powers.

Definition at line 937 of file TMultiDimFit.cxx.

◆ EvalError()

Double_t TMultiDimFit::EvalError ( const Double_t x,
const Double_t coeff = nullptr 
) const
virtual

Evaluate parameterization error at point x.

Optional argument coeff is a vector of coefficients for the parameterisation, fNCoefficients elements long.

Definition at line 904 of file TMultiDimFit.cxx.

◆ EvalFactor()

Double_t TMultiDimFit::EvalFactor ( Int_t  p,
Double_t  x 
) const
protectedvirtual

PRIVATE METHOD: Evaluate function with power p at variable value x.

Definition at line 952 of file TMultiDimFit.cxx.

◆ FindParameterization()

void TMultiDimFit::FindParameterization ( Option_t option = "")
virtual

Find the parameterization.

Options: None so far

For detailed description of what this entails, please refer to the class description

Definition at line 994 of file TMultiDimFit.cxx.

◆ Fit()

void TMultiDimFit::Fit ( Option_t option = "")
virtual

Try to fit the found parameterisation to the test sample.

Options M use Minuit to improve coefficients

Also, refer to class description

Definition at line 1013 of file TMultiDimFit.cxx.

◆ GetChi2()

Double_t TMultiDimFit::GetChi2 ( ) const
inline

Definition at line 140 of file TMultiDimFit.h.

◆ GetCoefficients()

const TVectorD * TMultiDimFit::GetCoefficients ( ) const
inline

Definition at line 142 of file TMultiDimFit.h.

◆ GetCoefficientsRMS()

const TVectorD * TMultiDimFit::GetCoefficientsRMS ( ) const
inline

Definition at line 143 of file TMultiDimFit.h.

◆ GetCorrelationMatrix()

const TMatrixD * TMultiDimFit::GetCorrelationMatrix ( ) const
inline

Definition at line 141 of file TMultiDimFit.h.

◆ GetError()

Double_t TMultiDimFit::GetError ( ) const
inline

Definition at line 144 of file TMultiDimFit.h.

◆ GetFunctionCodes()

Int_t * TMultiDimFit::GetFunctionCodes ( ) const
inline

Definition at line 145 of file TMultiDimFit.h.

◆ GetFunctions()

const TMatrixD * TMultiDimFit::GetFunctions ( ) const
inline

Definition at line 146 of file TMultiDimFit.h.

◆ GetHistograms()

virtual TList * TMultiDimFit::GetHistograms ( ) const
inlinevirtual

Definition at line 147 of file TMultiDimFit.h.

◆ GetMaxAngle()

Double_t TMultiDimFit::GetMaxAngle ( ) const
inline

Definition at line 148 of file TMultiDimFit.h.

◆ GetMaxFunctions()

Int_t TMultiDimFit::GetMaxFunctions ( ) const
inline

Definition at line 149 of file TMultiDimFit.h.

◆ GetMaxPowers()

Int_t * TMultiDimFit::GetMaxPowers ( ) const
inline

Definition at line 150 of file TMultiDimFit.h.

◆ GetMaxQuantity()

Double_t TMultiDimFit::GetMaxQuantity ( ) const
inline

Definition at line 151 of file TMultiDimFit.h.

◆ GetMaxStudy()

Int_t TMultiDimFit::GetMaxStudy ( ) const
inline

Definition at line 152 of file TMultiDimFit.h.

◆ GetMaxTerms()

Int_t TMultiDimFit::GetMaxTerms ( ) const
inline

Definition at line 153 of file TMultiDimFit.h.

◆ GetMaxVariables()

const TVectorD * TMultiDimFit::GetMaxVariables ( ) const
inline

Definition at line 154 of file TMultiDimFit.h.

◆ GetMeanQuantity()

Double_t TMultiDimFit::GetMeanQuantity ( ) const
inline

Definition at line 155 of file TMultiDimFit.h.

◆ GetMeanVariables()

const TVectorD * TMultiDimFit::GetMeanVariables ( ) const
inline

Definition at line 156 of file TMultiDimFit.h.

◆ GetMinAngle()

Double_t TMultiDimFit::GetMinAngle ( ) const
inline

Definition at line 157 of file TMultiDimFit.h.

◆ GetMinQuantity()

Double_t TMultiDimFit::GetMinQuantity ( ) const
inline

Definition at line 158 of file TMultiDimFit.h.

◆ GetMinRelativeError()

Double_t TMultiDimFit::GetMinRelativeError ( ) const
inline

Definition at line 159 of file TMultiDimFit.h.

◆ GetMinVariables()

const TVectorD * TMultiDimFit::GetMinVariables ( ) const
inline

Definition at line 160 of file TMultiDimFit.h.

◆ GetNCoefficients()

Int_t TMultiDimFit::GetNCoefficients ( ) const
inline

Definition at line 162 of file TMultiDimFit.h.

◆ GetNVariables()

Int_t TMultiDimFit::GetNVariables ( ) const
inline

Definition at line 161 of file TMultiDimFit.h.

◆ GetPolyType()

Int_t TMultiDimFit::GetPolyType ( ) const
inline

Definition at line 163 of file TMultiDimFit.h.

◆ GetPowerIndex()

Int_t * TMultiDimFit::GetPowerIndex ( ) const
inline

Definition at line 164 of file TMultiDimFit.h.

◆ GetPowerLimit()

Double_t TMultiDimFit::GetPowerLimit ( ) const
inline

Definition at line 165 of file TMultiDimFit.h.

◆ GetPowers()

const Int_t * TMultiDimFit::GetPowers ( ) const
inline

Definition at line 166 of file TMultiDimFit.h.

◆ GetPrecision()

Double_t TMultiDimFit::GetPrecision ( ) const
inline

Definition at line 167 of file TMultiDimFit.h.

◆ GetQuantity()

const TVectorD * TMultiDimFit::GetQuantity ( ) const
inline

Definition at line 168 of file TMultiDimFit.h.

◆ GetResidualMax()

Double_t TMultiDimFit::GetResidualMax ( ) const
inline

Definition at line 169 of file TMultiDimFit.h.

◆ GetResidualMaxRow()

Int_t TMultiDimFit::GetResidualMaxRow ( ) const
inline

Definition at line 171 of file TMultiDimFit.h.

◆ GetResidualMin()

Double_t TMultiDimFit::GetResidualMin ( ) const
inline

Definition at line 170 of file TMultiDimFit.h.

◆ GetResidualMinRow()

Int_t TMultiDimFit::GetResidualMinRow ( ) const
inline

Definition at line 172 of file TMultiDimFit.h.

◆ GetResidualSumSq()

Double_t TMultiDimFit::GetResidualSumSq ( ) const
inline

Definition at line 173 of file TMultiDimFit.h.

◆ GetRMS()

Double_t TMultiDimFit::GetRMS ( ) const
inline

Definition at line 174 of file TMultiDimFit.h.

◆ GetSampleSize()

Int_t TMultiDimFit::GetSampleSize ( ) const
inline

Definition at line 175 of file TMultiDimFit.h.

◆ GetSqError()

const TVectorD * TMultiDimFit::GetSqError ( ) const
inline

Definition at line 176 of file TMultiDimFit.h.

◆ GetSumSqAvgQuantity()

Double_t TMultiDimFit::GetSumSqAvgQuantity ( ) const
inline

Definition at line 177 of file TMultiDimFit.h.

◆ GetSumSqQuantity()

Double_t TMultiDimFit::GetSumSqQuantity ( ) const
inline

Definition at line 178 of file TMultiDimFit.h.

◆ GetTestError()

Double_t TMultiDimFit::GetTestError ( ) const
inline

Definition at line 179 of file TMultiDimFit.h.

◆ GetTestPrecision()

Double_t TMultiDimFit::GetTestPrecision ( ) const
inline

Definition at line 180 of file TMultiDimFit.h.

◆ GetTestQuantity()

const TVectorD * TMultiDimFit::GetTestQuantity ( ) const
inline

Definition at line 181 of file TMultiDimFit.h.

◆ GetTestSampleSize()

Int_t TMultiDimFit::GetTestSampleSize ( ) const
inline

Definition at line 182 of file TMultiDimFit.h.

◆ GetTestSqError()

const TVectorD * TMultiDimFit::GetTestSqError ( ) const
inline

Definition at line 183 of file TMultiDimFit.h.

◆ GetTestVariables()

const TVectorD * TMultiDimFit::GetTestVariables ( ) const
inline

Definition at line 184 of file TMultiDimFit.h.

◆ GetVariables()

const TVectorD * TMultiDimFit::GetVariables ( ) const
inline

Definition at line 185 of file TMultiDimFit.h.

◆ Instance()

TMultiDimFit * TMultiDimFit::Instance ( )
static

Return the static instance.

Definition at line 1099 of file TMultiDimFit.cxx.

◆ IsA()

TClass * TMultiDimFit::IsA ( ) const
inlineoverridevirtual
Returns
TClass describing current object

Reimplemented from TObject.

Definition at line 207 of file TMultiDimFit.h.

◆ IsFolder()

Bool_t TMultiDimFit::IsFolder ( ) const
inlineoverridevirtual

Returns kTRUE in case object contains browsable objects (like containers or lists of other objects).

Reimplemented from TObject.

Definition at line 188 of file TMultiDimFit.h.

◆ MakeCandidates()

void TMultiDimFit::MakeCandidates ( )
protectedvirtual

PRIVATE METHOD: Create list of candidate functions for the parameterisation.

See also class description

Definition at line 1110 of file TMultiDimFit.cxx.

◆ MakeChi2()

Double_t TMultiDimFit::MakeChi2 ( const Double_t coeff = nullptr)
virtual

Calculate Chi square over either the test sample.

The optional argument coeff is a vector of coefficients to use in the evaluation of the parameterisation. If coeff == 0, then the found coefficients is used. Used my MINUIT for fit (see TMultDimFit::Fit)

Definition at line 1240 of file TMultiDimFit.cxx.

◆ MakeCode()

void TMultiDimFit::MakeCode ( const char *  filename = "MDF",
Option_t option = "" 
)
virtual

Generate the file <filename> with .C appended if argument doesn't end in .cxx or .C.

The contains the implementation of the function:

Double_t <funcname>(Double_t *x)

which does the same as TMultiDimFit::Eval. Please refer to this method.

Further, the static variables:

Int_t    gNVariables
Int_t    gNCoefficients
Double_t gDMean
Double_t gXMean[]
Double_t gXMin[]
Double_t gXMax[]
Double_t gCoefficient[]
Int_t    gPower[]

are initialized. The only ROOT header file needed is Rtypes.h

See TMultiDimFit::MakeRealCode for a list of options

Definition at line 1290 of file TMultiDimFit.cxx.

◆ MakeCoefficientErrors()

void TMultiDimFit::MakeCoefficientErrors ( )
protectedvirtual

PRIVATE METHOD: Compute the errors on the coefficients.

For this to be done, the curvature matrix of the non-orthogonal functions, is computed.

Definition at line 1307 of file TMultiDimFit.cxx.

◆ MakeCoefficients()

void TMultiDimFit::MakeCoefficients ( )
protectedvirtual

PRIVATE METHOD: Invert the model matrix B, and compute final coefficients.

For a more thorough discussion of what this means, please refer to the class description

First we invert the lower triangle matrix fOrthCurvatureMatrix and store the inverted matrix in the upper triangle.

Definition at line 1365 of file TMultiDimFit.cxx.

◆ MakeCorrelation()

void TMultiDimFit::MakeCorrelation ( )
protectedvirtual

PRIVATE METHOD: Compute the correlation matrix.

Definition at line 1445 of file TMultiDimFit.cxx.

◆ MakeGramSchmidt()

Double_t TMultiDimFit::MakeGramSchmidt ( Int_t  function)
protectedvirtual

PRIVATE METHOD: Make Gram-Schmidt orthogonalisation.

The class description gives a thorough account of this algorithm, as well as references. Please refer to the class description

Definition at line 1504 of file TMultiDimFit.cxx.

◆ MakeHistograms()

void TMultiDimFit::MakeHistograms ( Option_t option = "A")
virtual

Make histograms of the result of the analysis.

This message should be sent after having read all data points, but before finding the parameterization

Options: A All the below X Original independent variables D Original dependent variables N Normalised independent variables S Shifted dependent variables R1 Residuals versus normalised independent variables R2 Residuals versus dependent variable R3 Residuals computed on training sample R4 Residuals computed on test sample

For a description of these quantities, refer to class description

Definition at line 1597 of file TMultiDimFit.cxx.

◆ MakeMethod()

void TMultiDimFit::MakeMethod ( const Char_t classname = "MDF",
Option_t option = "" 
)
virtual

Generate the file <classname>MDF.cxx which contains the implementation of the method:

Double_t <classname>::MDF(Double_t *x)

which does the same as TMultiDimFit::Eval. Please refer to this method.

Further, the public static members:

Int_t <classname>::fgNVariables
Int_t <classname>::fgNCoefficients
Double_t <classname>::fgDMean
Double_t <classname>::fgXMean[] //[fgNVariables]
Double_t <classname>::fgXMin[] //[fgNVariables]
Double_t <classname>::fgXMax[] //[fgNVariables]
Double_t <classname>::fgCoefficient[] //[fgNCoeffficents]
Int_t <classname>::fgPower[] //[fgNCoeffficents*fgNVariables]

are initialized, and assumed to exist. The class declaration is assumed to be in <classname>.h and assumed to be provided by the user.

See also
TMultiDimFit::MakeRealCode for a list of options

The minimal class definition is:

class <classname> {
public:
Int_t <classname>::fgNVariables; // Number of variables
Int_t <classname>::fgNCoefficients; // Number of terms
Double_t <classname>::fgDMean; // Mean from training sample
Double_t <classname>::fgXMean[]; // Mean from training sample
Double_t <classname>::fgXMin[]; // Min from training sample
Double_t <classname>::fgXMax[]; // Max from training sample
Double_t <classname>::fgCoefficient[]; // Coefficients
Int_t <classname>::fgPower[]; // Function powers
};
virtual Double_t Eval(const Double_t *x, const Double_t *coeff=nullptr) const
Evaluate parameterization at point x.
Double_t x[n]
Definition legend1.C:17

Whether the method <classname>::Eval should be static or not, is up to the user.

Definition at line 1744 of file TMultiDimFit.cxx.

◆ MakeNormalized()

void TMultiDimFit::MakeNormalized ( )
protectedvirtual

PRIVATE METHOD: Normalize data to the interval [-1;1].

This is needed for the classes method to work.

Definition at line 1756 of file TMultiDimFit.cxx.

◆ MakeParameterization()

void TMultiDimFit::MakeParameterization ( )
protectedvirtual

PRIVATE METHOD: Find the parameterization over the training sample.

A full account of the algorithm is given in the class description

Definition at line 1810 of file TMultiDimFit.cxx.

◆ MakeRealCode()

void TMultiDimFit::MakeRealCode ( const char *  filename,
const char *  classname,
Option_t option = "" 
)
protectedvirtual

PRIVATE METHOD: This is the method that actually generates the code for the evaluation the parameterization on some point.

It's called by TMultiDimFit::MakeCode and TMultiDimFit::MakeMethod.

The options are: NONE so far

Definition at line 1963 of file TMultiDimFit.cxx.

◆ Print()

void TMultiDimFit::Print ( Option_t option = "ps") const
overridevirtual

Print statistics etc.

Options are P Parameters S Statistics C Coefficients R Result of parameterisation F Result of fit K Correlation Matrix M Pretty print formula

Reimplemented from TObject.

Definition at line 2158 of file TMultiDimFit.cxx.

◆ Select()

Bool_t TMultiDimFit::Select ( const Int_t iv)
protectedvirtual

Selection method.

User can override this method for specialized selection of acceptable functions in fit. Default is to select all. This message is sent during the build-up of the function candidates table once for each set of powers in variables. Notice, that the argument array contains the powers PLUS ONE. For example, to De select the function f = x1^2 * x2^4 * x3^5, this method should return kFALSE if given the argument { 3, 4, 6 }

Definition at line 2365 of file TMultiDimFit.cxx.

◆ SetBinVarX()

void TMultiDimFit::SetBinVarX ( Int_t  nbbinvarx)
inline

Definition at line 195 of file TMultiDimFit.h.

◆ SetBinVarY()

void TMultiDimFit::SetBinVarY ( Int_t  nbbinvary)
inline

Definition at line 196 of file TMultiDimFit.h.

◆ SetMaxAngle()

void TMultiDimFit::SetMaxAngle ( Double_t  ang = 0)

Set the max angle (in degrees) between the initial data vector to be fitted, and the new candidate function to be included in the fit.

By default it is 0, which automatically chooses another selection criteria. See also class description

Definition at line 2377 of file TMultiDimFit.cxx.

◆ SetMaxFunctions()

void TMultiDimFit::SetMaxFunctions ( Int_t  n)
inline

Definition at line 198 of file TMultiDimFit.h.

◆ SetMaxPowers()

void TMultiDimFit::SetMaxPowers ( const Int_t powers)

Set the maximum power to be considered in the fit for each variable.

See also class description

Definition at line 2443 of file TMultiDimFit.cxx.

◆ SetMaxStudy()

void TMultiDimFit::SetMaxStudy ( Int_t  n)
inline

Definition at line 200 of file TMultiDimFit.h.

◆ SetMaxTerms()

void TMultiDimFit::SetMaxTerms ( Int_t  terms)
inline

Definition at line 201 of file TMultiDimFit.h.

◆ SetMinAngle()

void TMultiDimFit::SetMinAngle ( Double_t  ang = 1)

Set the min angle (in degrees) between a new candidate function and the subspace spanned by the previously accepted functions.

See also class description

Definition at line 2393 of file TMultiDimFit.cxx.

◆ SetMinRelativeError()

void TMultiDimFit::SetMinRelativeError ( Double_t  error)

Set the acceptable relative error for when sum of square residuals is considered minimized.

For a full account, refer to the class description

Definition at line 2458 of file TMultiDimFit.cxx.

◆ SetPowerLimit()

void TMultiDimFit::SetPowerLimit ( Double_t  limit = 1e-3)

Set the user parameter for the function selection.

The bigger the limit, the more functions are used. The meaning of this variable is defined in the class description

Definition at line 2433 of file TMultiDimFit.cxx.

◆ SetPowers()

void TMultiDimFit::SetPowers ( const Int_t powers,
Int_t  terms 
)
virtual

Define a user function.

The input array must be of the form (p11, ..., p1N, ... ,pL1, ..., pLN) Where N is the dimension of the data sample, L is the number of terms (given in terms) and the first number, labels the term, the second the variable. More information is given in the class description

Definition at line 2413 of file TMultiDimFit.cxx.

◆ Streamer()

void TMultiDimFit::Streamer ( TBuffer R__b)
overridevirtual

Stream an object of class TObject.

Reimplemented from TObject.

◆ StreamerNVirtual()

void TMultiDimFit::StreamerNVirtual ( TBuffer ClassDef_StreamerNVirtual_b)
inline

Definition at line 207 of file TMultiDimFit.h.

◆ TestFunction()

Bool_t TMultiDimFit::TestFunction ( Double_t  squareResidual,
Double_t  dResidur 
)
protectedvirtual

PRIVATE METHOD: Test whether the currently considered function contributes to the fit.

See also class description

Definition at line 2470 of file TMultiDimFit.cxx.

Member Data Documentation

◆ fBinVarX

Int_t TMultiDimFit::fBinVarX
protected

Number of bin in independent variables.

Definition at line 98 of file TMultiDimFit.h.

◆ fBinVarY

Int_t TMultiDimFit::fBinVarY
protected

Number of bin in dependent variables.

Definition at line 99 of file TMultiDimFit.h.

◆ fChi2

Double_t TMultiDimFit::fChi2
protected

Chi square of fit.

Definition at line 85 of file TMultiDimFit.h.

◆ fCoefficients

TVectorD TMultiDimFit::fCoefficients
protected

Vector of the final coefficients.

Definition at line 82 of file TMultiDimFit.h.

◆ fCoefficientsRMS

TVectorD TMultiDimFit::fCoefficientsRMS
protected

Vector of RMS of coefficients.

Definition at line 83 of file TMultiDimFit.h.

◆ fCorrelationCoeff

Double_t TMultiDimFit::fCorrelationCoeff
protected

Multi Correlation coefficient.

Definition at line 92 of file TMultiDimFit.h.

◆ fCorrelationMatrix

TMatrixD TMultiDimFit::fCorrelationMatrix
protected

Correlation matrix.

Definition at line 93 of file TMultiDimFit.h.

◆ fError

Double_t TMultiDimFit::fError
protected

Error from parametrization.

Definition at line 88 of file TMultiDimFit.h.

◆ fFitter

TVirtualFitter* TMultiDimFit::fFitter
protected

Definition at line 101 of file TMultiDimFit.h.

◆ fFunctionCodes

Int_t* TMultiDimFit::fFunctionCodes
protected

[fMaxFunctions] acceptance code

Definition at line 60 of file TMultiDimFit.h.

◆ fFunctions

TMatrixD TMultiDimFit::fFunctions
protected

Functions evaluated over sample.

Definition at line 58 of file TMultiDimFit.h.

◆ fgInstance

TMultiDimFit * TMultiDimFit::fgInstance = nullptr
staticprivate

Definition at line 25 of file TMultiDimFit.h.

◆ fHistogramMask

Byte_t TMultiDimFit::fHistogramMask
protected

Bit pattern of histograms used.

Definition at line 97 of file TMultiDimFit.h.

◆ fHistograms

TList* TMultiDimFit::fHistograms
protected

List of histograms.

Definition at line 96 of file TMultiDimFit.h.

◆ fIsUserFunction

Bool_t TMultiDimFit::fIsUserFunction
protected

Flag for user defined function.

Definition at line 105 of file TMultiDimFit.h.

◆ fIsVerbose

Bool_t TMultiDimFit::fIsVerbose
protected

Definition at line 106 of file TMultiDimFit.h.

◆ fMaxAngle

Double_t TMultiDimFit::fMaxAngle
protected

Max angle for accepting new function.

Definition at line 51 of file TMultiDimFit.h.

◆ fMaxFuncNV

Int_t TMultiDimFit::fMaxFuncNV
protected

fMaxFunctions*fNVariables

Definition at line 62 of file TMultiDimFit.h.

◆ fMaxFunctions

Int_t TMultiDimFit::fMaxFunctions
protected

max number of functions

Definition at line 59 of file TMultiDimFit.h.

◆ fMaxPowers

Int_t* TMultiDimFit::fMaxPowers
protected

[fNVariables] maximum powers

Definition at line 54 of file TMultiDimFit.h.

◆ fMaxPowersFinal

Int_t* TMultiDimFit::fMaxPowersFinal
protected

[fNVariables] maximum powers from fit;

Definition at line 68 of file TMultiDimFit.h.

◆ fMaxQuantity

Double_t TMultiDimFit::fMaxQuantity
protected

Max value of dependent quantity.

Definition at line 31 of file TMultiDimFit.h.

◆ fMaxResidual

Double_t TMultiDimFit::fMaxResidual
protected

Max residual value.

Definition at line 73 of file TMultiDimFit.h.

◆ fMaxResidualRow

Int_t TMultiDimFit::fMaxResidualRow
protected

Row giving max residual.

Definition at line 75 of file TMultiDimFit.h.

◆ fMaxStudy

Int_t TMultiDimFit::fMaxStudy
protected

max functions to study

Definition at line 61 of file TMultiDimFit.h.

◆ fMaxTerms

Int_t TMultiDimFit::fMaxTerms
protected

Max terms expected in final expr.

Definition at line 52 of file TMultiDimFit.h.

◆ fMaxVariables

TVectorD TMultiDimFit::fMaxVariables
protected

max value of independent variables

Definition at line 39 of file TMultiDimFit.h.

◆ fMeanQuantity

Double_t TMultiDimFit::fMeanQuantity
protected

Mean of dependent quantity.

Definition at line 30 of file TMultiDimFit.h.

◆ fMeanVariables

TVectorD TMultiDimFit::fMeanVariables
protected

mean value of independent variables

Definition at line 38 of file TMultiDimFit.h.

◆ fMinAngle

Double_t TMultiDimFit::fMinAngle
protected

Min angle for accepting new function.

Definition at line 50 of file TMultiDimFit.h.

◆ fMinQuantity

Double_t TMultiDimFit::fMinQuantity
protected

Min value of dependent quantity.

Definition at line 32 of file TMultiDimFit.h.

◆ fMinRelativeError

Double_t TMultiDimFit::fMinRelativeError
protected

Min relative error accepted.

Definition at line 53 of file TMultiDimFit.h.

◆ fMinResidual

Double_t TMultiDimFit::fMinResidual
protected

Min residual value.

Definition at line 74 of file TMultiDimFit.h.

◆ fMinResidualRow

Int_t TMultiDimFit::fMinResidualRow
protected

Row giving min residual.

Definition at line 76 of file TMultiDimFit.h.

◆ fMinVariables

TVectorD TMultiDimFit::fMinVariables
protected

min value of independent variables

Definition at line 40 of file TMultiDimFit.h.

◆ fNCoefficients

Int_t TMultiDimFit::fNCoefficients
protected

Dimension of model coefficients.

Definition at line 79 of file TMultiDimFit.h.

◆ fNVariables

Int_t TMultiDimFit::fNVariables
protected

Number of independent variables.

Definition at line 37 of file TMultiDimFit.h.

◆ fOrthCoefficients

TVectorD TMultiDimFit::fOrthCoefficients
protected

The model coefficients.

Definition at line 80 of file TMultiDimFit.h.

◆ fOrthCurvatureMatrix

TMatrixD TMultiDimFit::fOrthCurvatureMatrix
protected

Model matrix.

Definition at line 81 of file TMultiDimFit.h.

◆ fOrthFunctionNorms

TVectorD TMultiDimFit::fOrthFunctionNorms
protected

Norm of the evaluated functions.

Definition at line 65 of file TMultiDimFit.h.

◆ fOrthFunctions

TMatrixD TMultiDimFit::fOrthFunctions
protected

As above, but orthogonalised.

Definition at line 64 of file TMultiDimFit.h.

◆ fParameterisationCode

Int_t TMultiDimFit::fParameterisationCode
protected

Exit code of parameterisation.

Definition at line 86 of file TMultiDimFit.h.

◆ fPolyType

EMDFPolyType TMultiDimFit::fPolyType
protected

Fit object (MINUIT)

Type of polynomials to use

Definition at line 103 of file TMultiDimFit.h.

◆ fPowerIndex

Int_t* TMultiDimFit::fPowerIndex
protected

[fMaxTerms] Index of accepted powers

Definition at line 70 of file TMultiDimFit.h.

◆ fPowerLimit

Double_t TMultiDimFit::fPowerLimit
protected

Control parameter.

Definition at line 55 of file TMultiDimFit.h.

◆ fPowers

Int_t* TMultiDimFit::fPowers
protected

[fMaxFuncNV] where fMaxFuncNV = fMaxFunctions*fNVariables

Definition at line 69 of file TMultiDimFit.h.

◆ fPrecision

Double_t TMultiDimFit::fPrecision
protected

Relative precision of param.

Definition at line 90 of file TMultiDimFit.h.

◆ fQuantity

TVectorD TMultiDimFit::fQuantity
protected

Training sample, dependent quantity.

Definition at line 28 of file TMultiDimFit.h.

◆ fResiduals

TVectorD TMultiDimFit::fResiduals
protected

Vector of the final residuals.

Definition at line 72 of file TMultiDimFit.h.

◆ fRMS

Double_t TMultiDimFit::fRMS
protected

Root mean square of fit.

Definition at line 84 of file TMultiDimFit.h.

◆ fSampleSize

Int_t TMultiDimFit::fSampleSize
protected

Size of training sample.

Definition at line 42 of file TMultiDimFit.h.

◆ fShowCorrelation

Bool_t TMultiDimFit::fShowCorrelation
protected

print correlation matrix

Definition at line 104 of file TMultiDimFit.h.

◆ fSqError

TVectorD TMultiDimFit::fSqError
protected

Training sample, error in quantity.

Definition at line 29 of file TMultiDimFit.h.

◆ fSumSqAvgQuantity

Double_t TMultiDimFit::fSumSqAvgQuantity
protected

Sum of squares away from mean.

Definition at line 34 of file TMultiDimFit.h.

◆ fSumSqQuantity

Double_t TMultiDimFit::fSumSqQuantity
protected

SumSquare of dependent quantity.

Definition at line 33 of file TMultiDimFit.h.

◆ fSumSqResidual

Double_t TMultiDimFit::fSumSqResidual
protected

Sum of Square residuals.

Definition at line 77 of file TMultiDimFit.h.

◆ fTestCorrelationCoeff

Double_t TMultiDimFit::fTestCorrelationCoeff
protected

Multi Correlation coefficient.

Definition at line 94 of file TMultiDimFit.h.

◆ fTestError

Double_t TMultiDimFit::fTestError
protected

Error from test.

Definition at line 89 of file TMultiDimFit.h.

◆ fTestPrecision

Double_t TMultiDimFit::fTestPrecision
protected

Relative precision of test.

Definition at line 91 of file TMultiDimFit.h.

◆ fTestQuantity

TVectorD TMultiDimFit::fTestQuantity
protected

Test sample, dependent quantity.

Definition at line 44 of file TMultiDimFit.h.

◆ fTestSampleSize

Int_t TMultiDimFit::fTestSampleSize
protected

Size of test sample.

Definition at line 48 of file TMultiDimFit.h.

◆ fTestSqError

TVectorD TMultiDimFit::fTestSqError
protected

Test sample, Error in quantity.

Definition at line 45 of file TMultiDimFit.h.

◆ fTestVariables

TVectorD TMultiDimFit::fTestVariables
protected

Test sample, independent variables.

Definition at line 46 of file TMultiDimFit.h.

◆ fVariables

TVectorD TMultiDimFit::fVariables
protected

Training sample, independent variables.

Definition at line 36 of file TMultiDimFit.h.

Libraries for TMultiDimFit:

The documentation for this class was generated from the following files: