Very useful tools!
I found there is a small shift of the center position between the image I get from unagi (upper) and from hsc_map directly (lower):


They are fed with the same coordinate and the real size of the image is the same (3.6" x 3.6"). What could be a possible reason for this?
Thanks!

One easy way to maintain the version of the package is to use git tags or releases along with the setuptools_scm module. Check this out: https://pypi.org/project/setuptools-scm/

We can also setup a Travis CI automated build to automatically push new releases to PyPi

Then all you need to do to update the version is create a new tag on github, everything else is taken care of.

I've set it up for a number of packages before, I can take of it :-)

I like the default unagi plotting style but unagi may be a little too aggressive in its setting of the rc plot parameters. Simply doing import unagi sets the matloplib environment, and in particular sets plt.rc('text', usetex=True) here:

I'm runing into an issue when trying to run a code that imports unagi internally, in an environment that doesn't have LaTeX installed. My code crashes with:

FileNotFoundError: [Errno 2] No such file or directory: 'latex': 'latex'

(Why not just installing LaTeX? Because I'm running on Google Colab, I have to reinstall the environment every time the notebook is restarted, and installing LaTeX takes a lot of time....)

Here are 2 suggestions:

• Not importing plotting from __init__.py, it's not really necessary I don't think...
• "Hide" the plotting settings in a unagi.set() function, in the same way that seaborn does: https://seaborn.pydata.org/generated/seaborn.set.html

Thoughts?

Deal with HSC fluxes from image and catalog.

• Convert HSC images unit to fluxes in CGS unit, AB magnitude, and nanomaggies.
• Convert HSC catalog fluxes to AB magnitude, image unit, or naomaggies.
• Deal with pre-s18a/PDR1 catalog and after-s18a/PDR2` catalogs separately.

Update the config.py for S18A data release

Modified based on the script provided by NAOJ people

TODO:

• Modify the NAOJ script to generate the picture
• Generate 3-color figure using Lupton SDSS true color
• Option to add scale bar or compass
• Write test function

Documents, documents, documents...

I was looking through the hsc-das-cutout documentation and noticed they have a bulk upload api for space delimited tables (https://hsc-release.mtk.nao.ac.jp/das_cutout/pdr2/manual.html#list-to-upload)

I think this might be a useful addition to your already great project.
Here is an example table from the documentation:

#? rerun      filter    ra       dec       sw     sh  # column descriptor
   pdr2_wide  HSC-G  -1:36:00  00:00:00  2asec  2asec #
   pdr2_wide  HSC-R  -1:36:00  00:00:00  2asec  2asec #
   pdr2_wide  HSC-I  -1:36:00  00:00:00  2asec  2asec # list of coordinates
   pdr2_wide  HSC-Z  -1:36:00  00:00:00  2asec  2asec #
   pdr2_wide  HSC-Y  -1:36:00  00:00:00  2asec  2asec #

Assuming the table above was in a file called sources.txt we could pull them all at the same time with the following code using the requests library:

import requests
from tqdm import tqdm

creds = {
    "user":"your-username-here",
    "secret":"your-password-here"
}

with open("sources.txt", "r") as f:
    # https://stackoverflow.com/a/37573701
    r = requests.post("https://hsc-release.mtk.nao.ac.jp/das_cutout/pdr2/cgi-bin/cutout",
                      files={"list":f}, auth=(creds["user"], creds["secret"]), stream=True)

block_size = 1024
with open("downloaded.tar.gz", "wb") as f:
    for data in tqdm(r.iter_content(block_size), unit='KB', desc=f"Downloading file"):
        f.write(data)

Unagi is awesome, let's make it trivially installable with pip by pushing it to PyPi ;-)

@dr-guangtou Happy to help with this, let me know if I can provide any assistance.

Issue found by Connor Bottrell:

I find that the SQL query tool in unagi doesn’t recognize the s20a_wide.photoz_mizuki table:

import unagi
dr = unagi.hsc.Hsc(dr='dr3', rerun='s20a_wide')
sql_cmd = [
    'SELECT f.object_id, f.ra, f.dec, z.photoz_median',
    'FROM s20a_wide.forced as f',
    'LEFT JOIN s20a_wide.photoz_mizuki as z',
    'ON f.object_id = z.object_id',
    'LIMIT 10'
          ]
sql_cmd = ' '.join(sql_cmd)
print(sql_cmd)
result = dr.sql_query(sql_cmd, from_file=False, verbose=True)

Which gives:

Waiting for query to finish... [Failed]
query error: - Query Error:
relation "s20a_wide.photoz_mizuki" does not exist
LINE 1: ..., z.photoz_median FROM s20a_wide.forced LEFT JOIN s20a_wide....

Add a filters.py module to deal with HSC filters:

• Convert the HSC .txt format filter response curves to the .par format used by kcorrect and sedpy
• Get the center wavelength and effective wavelength of each filter
• Easy to access HSC filter list for both Wide filters and all fitlers
• Plot the filter response curves.

Hi, I noticed that nb1010 filter is supported in hsc.py but not in filters.py. Is it possible to add filter information retrieval for nb1010 in filters.py as well? Thanks

• Function that convert from 2nd-moment to traditional size, shape, and positional angle.
• Vice versa

https://hscdata.mtk.nao.ac.jp/hsc_ssp/dr4/s21a/index.html

• There are multiple ways to do this using SQL search, but neither is fast enough to check a large number of galaxies.
• If we could keep a copy of the Tract/Patch information of HSC footprints, there might be faster way to do this.

Will think more on this...