JSOC/proj/sharp/apps/errorprop.c - diff

Return to errorprop.c CVS log

Up to [Development] / JSOC / proj / sharp / apps

version 1.1, 2012/08/23 08:38:33

version 1.2, 2019/02/02 04:53:58

Line 32 int errorprop (float *bTotal, float *bAz

// Xudong Oct 18 2011: Fill factor variances covariances are all NaNs. removed

// NNB interpolation, just 1 point

// Fixed definition of azi for all derivatives

// Feb 1 2019, dBt/dXX is actually dBy/dXX, so the sign is wrong; however they are used in pairs (squared) so final result was okay

static double raddeg = M_PI / 180.;

double b, inc, azim;

Line 90 int errorprop (float *bTotal, float *bAz

Line 91 int errorprop (float *bTotal, float *bAz

dBpdInc = b * (- a11 * cos(inc) * sin(azim) + a12 * cos(inc) * cos(azim) - a13 * sin(inc));

dBpdAzim = b * (- a11 * sin(inc) * cos(azim) - a12 * sin(inc) * sin(azim));

dBtdBtotal = (- a21 * sin(inc) * sin(azim) + a22 * sin(inc) * cos(azim) + a23 * cos(inc));

dBtdBtotal = (- a21 * sin(inc) * sin(azim) + a22 * sin(inc) * cos(azim) + a23 * cos(inc)) * (-1);

dBtdInc = b * (- a21 * cos(inc) * sin(azim) + a22 * cos(inc) * cos(azim) - a23 * sin(inc));

dBtdInc = b * (- a21 * cos(inc) * sin(azim) + a22 * cos(inc) * cos(azim) - a23 * sin(inc)) * (-1);

dBtdAzim = b * (- a21 * sin(inc) * cos(azim) - a22 * sin(inc) * sin(azim));

dBtdAzim = b * (- a21 * sin(inc) * cos(azim) - a22 * sin(inc) * sin(azim)) * (-1);

dBrdBtotal = (- a31 * sin(inc) * sin(azim) + a32 * sin(inc) * cos(azim) + a33 * cos(inc));

dBrdInc = b * (- a31 * cos(inc) * sin(azim) + a32 * cos(inc) * cos(azim) - a33 * sin(inc));

Legend: