EMAN2
Protected Member Functions
EMAN::ModelHelixProcessor Class Reference

#include <processor.h>

Inheritance diagram for EMAN::ModelHelixProcessor:
Inheritance graph
[legend]
Collaboration diagram for EMAN::ModelHelixProcessor:
Collaboration graph
[legend]

List of all members.

Protected Member Functions

float radprofile (float r, int type)

Detailed Description

Definition at line 7344 of file processor.h.


Member Function Documentation

float ModelHelixProcessor::radprofile ( float  r,
int  type 
) [protected]

Definition at line 10446 of file processor.cpp.

{
        // r in angstrom
        double ret = 0;//iterations ==> rounding is problematic with float types if 15 < r < 20 and really bad if 20 < r < 30
        if (type == 0) { // pure Gaussian falloff
                r /= 2;
                ret = exp(-r * r);
        } else if (type == 1) { // pure Gaussian falloff + negative dip, so mean is 0
                r /= 2;
                ret = (1 - r * r / 4) * exp(-r * r / 4);
        } else if (type == 2) {
                // polynomial fitting to the radial profile of real helix density
                // f=a0*x^n+a1+x^(n-1)+ ... +a[n-1]*x+an
                //float an[11]={2.847024584977009e-10,-3.063997224364090e-08,1.418801040660860e-06,-3.678676414383996e-05,5.804871622801710e-04,-5.640340018430164e-03,3.208802421493864e-02,-9.068475313823952e-02,7.097329559749284e-02,-9.993347339658298e-02,1.000000000000000e+00};

                // now the fitting to the original profile
                if (r >= 12.2)
                        return 0; //We don't want that part of the polynomial --> goes way below zero
                static float an[15] = { -3.9185246832229140e-16f,
                                3.3957205298900993e-14f, 2.0343351971222658e-12f,
                                -4.4935965816879751e-10f, 3.0668169835080933e-08f,
                                -1.1904544689091790e-06f, 2.9753088549414953e-05f,
                                -4.9802112876220150e-04f, 5.5900917825309360e-03f,
                                -4.0823714462925299e-02f, 1.8021733669148599e-01f,
                                -4.0992557296268717e-01f, 3.3980328566901458e-01f,
                                -3.6062024812411908e-01f, 1.0000000000000000e+00f };

                ret = an[0];
                for (int i = 1; i < 15; i++) {
                        ret = ret * r + an[i];
                }
        }
        return (float)ret;
}

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