reikonakajima / lensing Goto Github PK

Because I need to compare my results to TreeCorr. It may also come in handy in the future.

The weights on the lens objects (due to lensing weights) must be similar to the object weights used for galaxy correlation function.
Add this feature to the gglens calculation code (should be done in GGLens.h/cpp) such that it spits out the lensing weight for each lens, for the galaxy correlation code to use.

It's probably the randoms subtraction that is setting the binned data to 0.

Check the n(z) output compared to the Python output.
It's possibly because of the float in the C++ code, compared to the double used in Python.

To make the code run faster. Right now, it's taking forever.
(Also applies for gglens_random_kids.cpp)

Implement flags for the c++ code arguments for gglens_gama_kids and gglens_random_kids.

Given a c-term correction, apply to the KiDS catalog when loading the data.

Add a driver code "gglens_starhalo.cpp".

Add gglens_gama_kids.cpp code for calculating gglens signal for GAMA with KIDS sources.

Use tabulated angular gglens randoms data to subtract the randoms from the true GAMA signal (in gglens_gama_kids.cpp).

At least have the minimum info on the mean separation R.

The current code is such that the spherical trigonometry angles to calculate (1) angular separation and (2) angle beta to convert the shear into tangential shear for gg-lens, are determined from a spherical configuration from the outside of the sphere.

Since astronomical observations happen from the inside of the sphere (observer being in the center), and the shears are defined in a coordinate system projected into the inside of the sphere, we need to change the spherical trig angles definition, specifically for angle beta.

Making this change should fix the following, as addressed in issue #12:
(a) the g2 flip would no longer be required in the KiDS shear catalog
(b) the cross shear will have the correct sign (i.e. will agree with TreeCorr and Edo's code)