This is the anisotropy methods study by M. Stolpovskiy
Work started in Dec 2018
tools.py : a set of useful PYTHON tools for anisotropy, like:
- Simulation of observations
- Reconstruction of the dipole
- Setting the dipole upper limit
- Tophat smoothing
- Sky filtering
- Anisotropy significance
I use the Jupyter Notebook a lot. Here are some notebooks created:
significance_formula.ipynb :
discussion about the significance formula for the tophat smoothed maps
dipole_sensitivity.ipynb :
The most general way to get the UL for the dipole amplitude.
The experimental UL discussed.
dipole_sensitivity_thr.ipynb :
More optimized way, with noise thresholding. Allows to get lower UL.
rate-based_MC.ipynb :
MC comparison of the event-shuffling and rate-based methods.
Conclusion: rate-based is good and event-shuffling is bad
However, the UL results are still valid, because the
badness of the event-shuffling is only evident with
high enough dipole amplitude.
In order to run any of these notebooks you'll need an iso_map.npy -- a file
which contains a numpy array with the isotropy (exposure) map in galactic coordinates.
Software requirements: Python 2 with
- numpy
- matplotlib
- scipy
- healpy of version