import numpy as np
from .baseMetric import BaseMetric
import lsst.sims.maf.utils as mafUtils
import lsst.sims.utils as utils
__all__ = ['DcrPrecisionMetric']
[docs]class DcrPrecisionMetric(BaseMetric):
"""Determine how precise a DCR correction could be made
Parameters
----------
atm_err : float
Minimum error in photometry centroids introduced by the atmosphere (arcseconds). Default 0.01.
"""
def __init__(self, metricName='DCRprecision', seeingCol='seeingFwhmGeom',
m5Col='fiveSigmaDepth', HACol='HA', PACol='paraAngle',
filterCol='filter', atm_err=0.01, SedTemplate='flat',
rmag=20., **kwargs):
self.m5Col = m5Col
self.filterCol = filterCol
self.PACol = PACol
self.seeingCol = seeingCol
self.mags = {}
self.filters = ['u', 'g', 'r', 'i', 'z', 'y']
if SedTemplate == 'flat':
for f in self.filters:
self.mags[f] = rmag
else:
self.mags = utils.stellarMags(SedTemplate, rmag=rmag)
cols = ['ra_dcr_amp', 'dec_dcr_amp', seeingCol, m5Col, filterCol, 'zenithDistance', PACol]
units = 'arcseconds'
self.atm_err = atm_err
super(DcrPrecisionMetric, self).__init__(cols, metricName=metricName, units=units,
**kwargs)
[docs] def run(self, dataSlice, slicePoint=None):
snr = np.zeros(len(dataSlice), dtype='float')
for filt in self.filters:
inFilt = np.where(dataSlice[self.filterCol] == filt)
snr[inFilt] = mafUtils.m52snr(self.mags[filt], dataSlice[self.m5Col][inFilt])
position_errors = np.sqrt(mafUtils.astrom_precision(dataSlice[self.seeingCol], snr)**2 +
self.atm_err**2)
x_coord = np.tan(np.radians(dataSlice['zenithDistance']))*np.sin(np.radians(dataSlice[self.PACol]))
x_coord2 = np.tan(np.radians(dataSlice['zenithDistance']))*np.cos(np.radians(dataSlice[self.PACol]))
# Things should be the same for RA and dec.
# Now I want to compute the error if I interpolate/extrapolate to +/-1.
# function is of form, y=ax. a=y/x. da = dy/x.
# Only strictly true if we know the unshifted position. But this should be a reasonable approx.
slope_uncerts = position_errors/x_coord
slope_uncerts2 = position_errors/x_coord2
total_slope_uncert = 1./np.sqrt(np.sum(1./slope_uncerts**2)+np.sum(1./slope_uncerts2**2))
# So, this will be the uncertainty in the RA or Dec offset at x= +/- 1. A.K.A., the uncertainty in the slope
# of the line made by tan(zd)*sin(PA) vs RA offset
# or the line tan(zd)*cos(PA) vs Dec offset
# Assuming we know the unshfted position of the object (or there's little covariance if we are fitting for both)
result = total_slope_uncert
return result