IPP Software Navigation Tools IPP Links Communication Pan-STARRS Links

Ignore:
Timestamp:
Jun 26, 2012, 11:25:02 AM (14 years ago)
Author:
eugene
Message:

fix numerical round-off errors and aperture bias in ppSimInsertStars

File:
1 edited

Legend:

Unmodified
Added
Removed
  • trunk/ppSim/src/ppSimInsertGalaxies.c

    r32350 r34083  
    22static char *defaultModel = "PS_MODEL_SERSIC";
    33
    4 float imageSum (psImage *image) {
    5     float sum = 0.0;
     4// this is used to measure the inserted flux as a difference of 2 large numbers : must use
     5// 'double' or we get floating point errors and trends on a scale of ~0.005 mags at > -10 mags
     6// instrumental
     7double imageSum (psImage *image) {
     8    double sum = 0.0;
    69    for (int iy = 0; iy < image->numRows; iy++) {
    710        for (int ix = 0; ix < image->numCols; ix++) {
     
    165168       
    166169        // insert the source flux in the image
    167         float sum1 = imageSum(source->pixels);
     170        double sum1 = imageSum(source->pixels);
    168171        pmSourceAddWithOffset (source, PM_MODEL_OP_FULL, 0xff, dX, dY);
    169         float sum2 = imageSum(source->pixels);
     172        double sum2 = imageSum(source->pixels);
    170173        float flux = sum2 - sum1;
    171174
Note: See TracChangeset for help on using the changeset viewer.