pisa
pisa copied to clipboard
icecube
Monte Carlo-based data analysis
Making the bin predictions differentiable was suggested multiple times now. That would for example help minimizers and allow to extract the Fisher-information at the best-fit point. This will be a...
We use `from scipy.interpolate import interp1d` in multiple parts of the code, namely - [https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/core/prior.py#L14](https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/core/prior.py#L14) - [https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/stages/background/atm_muons.py#L7](https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/stages/background/atm_muons.py#L7) - [https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/utils/spline_smooth.py#L10](https://github.com/icecube/pisa/blob/cfeb9b45a38f981059197c293f1a53aa5ab81a39/pisa/utils/spline_smooth.py#L10) - The new stage introduced in #https://github.com/icecube/pisa/pull/798 At the moment we...
Global best-fit values from various [NuFIT](http://www.nu-fit.org) versions beyond 2.2 are defined in the config files in https://github.com/icecube/pisa/tree/f739ae4280ebdd5dc36d5b5ed0132f1feec34368/pisa_examples/resources/settings/osc, but the readme doesn't reflect that. It might also be helpful to point...
The README in the [pipeline examples](https://github.com/icecube/pisa/tree/master/pisa_examples/resources/settings/pipeline) seems to be super out of date. Maybe someone could clean this up a bit.
In [these lines](https://github.com/icecube/pisa/blob/15eed8557db13638ff241cfbaee8a5ca779f175a/pisa/stages/discr_sys/ultrasurfaces.py#L136-L137), the US fit results are loaded, but this expects an absolute path (so not generic for different clusters/wgs. This could simply be replaced with the [find_resource](https://github.com/icecube/pisa/blob/master/pisa/utils/resources.py#L38) function...
Apart from the issues that * supposedly, "the analysis guide" explains the difference between "parameterized-Monte Carlo (MC) stages and reweighted-MC stages", * and that the readme refers to the ezFIT...
* https://github.com/icecube/pisa/blob/master/pisa_tests/test_changes_with_combined_pidreco.py uses nonexistent pipeline configs to address pisa 2 vs. 3 compatibility issues * https://github.com/icecube/pisa/blob/master/pisa_tests/test_command_lines.sh tries to run a nonexistent python script (pisa/scripts/make_toy_events.py) and assumes every `.py` file to...
This should be a part of general [PISA conventions](https://github.com/icecube/pisa/blob/master/pisa/general_conventions.md) (again). The short readme https://github.com/icecube/pisa/blob/master/pisa/stages/README.md contains some useful bits and pieces but is far from comprehensive. See also https://github.com/icecube/pisa/issues/513.
I decided to open this issue as a place to document any future discussion about this topic. The PISA guide on general conventions has been stating for a long time...
Several services implement numpy generalised universal functions using `from numba import guvectorize`. Unless I'm misunderstanding, `numba.set_num_threads(PISA_NUM_THREADS)` in https://github.com/icecube/pisa/blob/master/pisa/utils/numba_tools.py#L71 is always executed (for the parallel target) only after some involved chain...