astro_toot_colors.py

from astro_utils import *
from astropy.time import Time
from mastodon_bot import connect_bot
from astro_list_ngc import list_ngc, ngc_html, choose_fits
from glob import glob
from skimage.transform import resize
from pyongc import ongc
##
df_prev = pd.read_csv('ngc.csv', sep=',')
df = list_ngc()
df.to_csv('ngc.csv', sep=',', index=False)
ngc_html()
masto, loc = connect_bot()
##
# if (df.iloc[0]['collected_to'] <= df_prev.iloc[0]['collected_to']) and (loc == 'github'):
#     print('no new NGC')
# else:
if loc == 'local':
    row = 0
    pkl = True
else:
    row = 0
    pkl = False
tgt = df['target_name'][row]
log_csv = f'../../logs/{tgt}_{df["collected_from"][row][:10]}.csv'
t_min = [np.floor(Time(df_prev['collected_from'][row]).mjd),
         np.ceil(Time(df_prev['collected_to'][row]).mjd)]
args = {'target_name': tgt,
        't_min': t_min,
        'obs_collection': "JWST",
        'calib_level': 3,
        'dataRights': 'public',
        'intentType': 'science',
        'dataproduct_type': "image"}
table = Observations.query_criteria(**args)
if not os.path.isdir('data'):
    os.system('mkdir data')
if not os.path.isdir('data/'+tgt):
    os.system('mkdir data/'+tgt)
os.chdir('data/'+tgt)
# check if pkl was saved locally
download_fits_files(list(table['dataURL']))
## make image
# read the files and for each filter, choose smaller and close to target images
chosen_df = choose_fits()
chosen_df.to_csv(log_csv, index=False, sep=',')
use = chosen_df['chosen'].to_numpy()
files = np.asarray(chosen_df['file'])
todelete = files[~use]
for rm in todelete:
    os.system('rm '+rm)
# see if we have both MIRI and NIRCam, choose RGB method accordingly
files = files[use]
filt = filt_num(files)
files = files[np.argsort(filt)]
filt = np.sort(filt)
mn = np.zeros((len(files),2), bool)
for ii in range(len(files)):
    if 'miri' in files[ii]:
        mn[ii,0] = True
    if 'nircam' in files[ii]:
        mn[ii,1] = True
    if mn[ii, :].sum() == 0:
        raise Exception('no miri and no nircam')
    elif mn[ii, :].sum() == 2:
        raise Exception('both miri and nircam')

method = 'rrgggbb'
if np.mean(mn[:,0]) == 1:
    instrument = 'MIRI'
elif np.mean(mn[:,1]) == 1:
    instrument = 'NIRCam'
else:
    instrument = 'NIRCam + MIRI'
    method = 'mnn'
os.chdir('..')
# TODO: do something about duplicate filters, avoid error when one layer is not in overlap, decide if to use 0.5 1 1
auto_plot(tgt, exp=list(files), png=True, pow=[1, 1, 1], pkl=pkl, resize=True, method=method, plot=False)
##

what = '.'
if df['NGC'][row] > 0:
    obj = ongc.get('NGC' + str(df['NGC'][row]))
    if obj is not None:
        what = ', A ' + obj.type.lower() + ' in ' + obj.constellation + '.'
toot = tgt + what + ' An image of #JWST data released on ' + df_prev['collected_from'][row][:10] + \
       ', instrument: ' + instrument + ', filters: ' + str(filt.astype(int)) + '.\n' + \
       ' For the latest NGC images by JWST, see:\n https://yuval-harpaz.github.io/astro/ngc_grid.html'
img = plt.imread(tgt + '.png')
size = os.path.getsize(tgt+'.png')
mb2 = 2 * 10 ** 6  # mastodon allows 2MB
for decrease in [0, 0.01, 0.1, 0.2, 0.3, 0.4, 0.5]:
    try:
        if size >= mb2:
            ratio = mb2 / size - decrease
        else:
            ratio = 1 - decrease
        img = resize(img, (int(ratio ** 0.5 * img.shape[0]), int(ratio ** 0.5 * img.shape[1])))
        plt.imsave(tgt+'.png', img, cmap='gray')
        metadata = masto.media_post(tgt+'.png', "image/PNG")
        masto.status_post(toot, media_ids=metadata["id"])
        print('toot image')
        break
    except:
        print('fialed decrease of '+str(decrease))