An adaptation of the simple binner dwarf galaxy search algorithm for use on generic clusters for the MC 2021 dwarf search.

This version follows simple_adl which uses data pulled directly from the Astro Data Lab. Thanks Sidney and Will!

simple_adl originates from simple.

Clone this repository to the machine you wish to run the code from.

Run the following command (consider adding to your .bashrc/.zshrc/etc.):

export PYTHONPATH=<path to simple_adl>:$PYTHONPATH

where <path to simple_adl> is the path to the inner simple_adl repository (ie, the full string should end with */simple_adl:$PYTHONPATH).

Create the conda environment with

conda env create -f conda.yaml

This will create a conda environment called simple_adl that includes all of the necessary software.

Activate the conda environment through conda activate simple_adl. Note that this needs to be done from a directory that does not have a simple_adl folder.

Run init_simple.py to generate a default config.yaml configuration file. This can be edited as needed; in particular, you may want to change the profile entry.

Update search_coordinates.csv with the coordinates to be searched.

Open parallel_search.py. Enter the number of threads to run the search on in n_threads. On linux, use:

 $ lspcu 

to check how many threads per core your machine has (e.g., if there are 2 per core then a 7 core computer uses 14 threads).

NOTE: as the algorithm is subprocess intense, it is recommended to try a few test runs with less threads and find the best compromise.

Adapt the parallel_search.py filepaths to your machine.

Run on the command line:

python parallel_search.py 

Or, if on a VPN cluster try

nohup python parallel_search.py &

this keeps the process running in the background even if the connection breaks. The output is muted, so do

tail -f nohup.out

to watch the total output log being written.

Monitor the log_dir directory for output log information per search.

Once finished, run:

 python make_list.py

Inspect your candidate_list.csv.

Open parallel_plot_hotspot.py. You can either read in candidate_list.csv or, after applying any cuts (such as 5 sigma), read in the new candidate_list.csv file with a modified name (such as candidate_list_5sigma.csv) and keep your old one for reference. Change the file name accordingly in the code.

Set your number of threads and paths, run:

python parallel_plot_hotspot.py

And voila! Happy searching!

To check which log files are empty (incomplete run), run in the log directory:

grep -L "Job done" *

this prints log files without this checkpoint, suggestive of an error.

Check if there is no corresponding .csv file in the results. If there is, delete the .csv and run the algorithm again (tip: without nohup as to monitor the output live).

NOTE: If there is no 'Job done' in the log, but the above command says that there is, please view the log in a different editor. Sometimes the editor doesn't load the entire log.

To search for DELVE 2, run the following:

python search.py --ra 28.77 --dec -68.25 

or

python search.py --nside 32 --ipix 11812

To plot DELVE 2, run

python plot_hotspot.py --ra 28.77 --dec -68.25 --mod 19.26