def _get_data(sector, cdips_cat_vnum=None, is_not_cdips_still_good=0):
if is_not_cdips_still_good:
classifxn_csv = os.path.join(
CLASSIFXNDIR,
"sector-{}_PCs_NOT_CDIPS_STILL_GOOD.csv".format(sector))
else:
classifxn_csv = os.path.join(
CLASSIFXNDIR, "sector-{}_PCs_CLEAR_THRESHOLD.csv".format(sector))
df = pd.read_csv(classifxn_csv, sep=';')
cdips_df = get_cdips_catalog(ver=cdips_cat_vnum)
pfpath = os.path.join(
resultsbase, 'cdips_lc_periodfinding/'
'sector-{}/'.format(sector) +
'initial_period_finding_results_supplemented.csv')
pfdf = pd.read_csv(pfpath, sep=';')
supppath = os.path.join(
resultsbase, 'cdips_lc_stats/sector-{}/'.format(sector) +
'supplemented_cdips_lc_statistics.txt')
supplementstatsdf = pd.read_csv(supppath, sep=';')
toidf = get_toi_catalog()
ctoidf = get_exofop_ctoi_catalog()
return df, cdips_df, pfdf, supplementstatsdf, toidf, ctoidf
def _get_nbhd_dataframes():
"""
WARNING!: this "bright" subset is a crossmatch between the full NGC 2516
target list (CG18+KC19+M21), and the CDIPS target catalog (G_Rp<16; v0.4).
However, since the CDIPS targets didn't incorporate M21, it's not as direct
of a match as desired. This is fine for understanding the auto-detection of
rotation periods. But for overall cluster rotation period completeness,
it's not.
The "neighborhood" was selected via
bounds = { 'parallax_lower': 1.5, 'parallax_upper': 4.0, 'ra_lower': 108,
'ra_upper': 132, 'dec_lower': -76, 'dec_upper': -45 }
nbhd_df = query_neighborhood(bounds, groupname, n_max=6000,
overwrite=False, manual_gmag_limit=17)
This procedure yields:
Got 7052 neighbors with Rp<16
Got 893 in core from CDIPS target catalog
Got 1345 in corona from CDIPS target catalog
"""
df = get_cdips_catalog(ver=0.4)
nbhd_df, core_df, halo_df, full_df, target_df = _get_fullfaint_dataframes()
#
# do the "bright" selection by a crossmatch between the full target list
# and the CDIPS catalog. so implicitly, it's a CDIPS target star catalog
# match. this misses some Meingast stars, b/c they weren't in the CDIPS
# v0.4 target list. but this
#
cdips_df = df['source_id']
mdf = full_df.merge(cdips_df, on='source_id', how='inner')
nbhd_df = nbhd_df[nbhd_df.phot_rp_mean_mag < 16]
core_df = mdf[mdf.subcluster == 'core']
halo_df = mdf[mdf.subcluster == 'halo']
print(42 * '.')
print('"Bright" sample:')
print(f'...Got {len(nbhd_df)} neighbors with Rp<16')
print(f'...Got {len(core_df)} in core from CDIPS target catalog')
print(f'...Got {len(halo_df)} in corona from CDIPS target catalog')
print(42 * '.')
return nbhd_df, core_df, halo_df, full_df, target_df
def plot_TIC268_nbhd_small(outdir=RESULTSDIR):
set_style()
df = get_cdips_catalog(ver=0.4)
kc19_sel = ((df.cluster.str.contains('NGC_2516')) &
(df.reference.str.contains('Kounkel_2019')))
cg18_sel = ((df.cluster.str.contains('NGC_2516')) &
(df.reference.str.contains('CantatGaudin_2018')))
target_df = df[df.source_id == 5489726768531119616] # TIC 2683...
kc19_df = df[kc19_sel]
cg18_df = df[cg18_sel]
kc19_df = kc19_df[~(kc19_df.source_id.isin(cg18_df.source_id))]
##########
# NGC 2516 rough
bounds = {
'parallax_lower': 1.5,
'parallax_upper': 4.0,
'ra_lower': 108,
'ra_upper': 132,
'dec_lower': -76,
'dec_upper': -45
}
groupname = 'customngc2516'
nbhd_df = query_neighborhood(bounds,
groupname,
n_max=6000,
overwrite=False,
manual_gmag_limit=17)
sel_nbhd = ((~nbhd_df.source_id.isin(kc19_df.source_id))
& (~nbhd_df.source_id.isin(cg18_df.source_id)))
from copy import deepcopy
orig_nbhd_df = deepcopy(nbhd_df)
nbhd_df = nbhd_df[sel_nbhd]
print(f'Got {len(nbhd_df)} neighbors')
print(f'Got {len(cg18_df)} in core')
print(f'Got {len(kc19_df)} in corona')
##########
plt.close('all')
f, axs = plt.subplots(figsize=(4, 3), ncols=2)
xv, yv = 'ra', 'dec'
axs[0].scatter(nbhd_df[xv],
nbhd_df[yv],
c='gray',
alpha=0.5,
zorder=2,
s=7,
rasterized=True,
linewidths=0,
label='Field',
marker='.')
axs[0].scatter(kc19_df[xv],
kc19_df[yv],
c='lightskyblue',
alpha=0.9,
zorder=3,
s=7,
rasterized=True,
linewidths=0.15,
label='Corona',
marker='.',
edgecolors='k')
axs[0].scatter(cg18_df[xv],
cg18_df[yv],
c='k',
alpha=0.9,
zorder=4,
s=7,
rasterized=True,
label='Core',
marker='.')
axs[0].plot(target_df[xv],
target_df[yv],
alpha=1,
mew=0.5,
zorder=8,
label='TOI 1937',
markerfacecolor='lightskyblue',
markersize=14,
marker='*',
color='black',
lw=0)
axs[0].set_xlabel(r'$\alpha$ [deg]')
axs[0].set_ylabel(r'$\delta$ [deg]')
axs[0].set_xlim([108, 132])
axs[0].set_ylim([-76, -45])
##########
get_yval = (lambda _df: np.array(_df['phot_g_mean_mag'] + 5 * np.log10(_df[
'parallax'] / 1e3) + 5))
get_xval = (lambda _df: np.array(_df['phot_bp_mean_mag'] - _df[
'phot_rp_mean_mag']))
axs[1].scatter(get_xval(nbhd_df),
get_yval(nbhd_df),
c='gray',
alpha=0.8,
zorder=2,
s=7,
rasterized=True,
linewidths=0,
label='Field',
marker='.')
axs[1].scatter(get_xval(kc19_df),
get_yval(kc19_df),
c='lightskyblue',
alpha=1,
zorder=3,
s=7,
rasterized=True,
linewidths=0.15,
label='Corona',
marker='.',
edgecolors='k')
axs[1].scatter(get_xval(cg18_df),
get_yval(cg18_df),
c='k',
alpha=0.9,
zorder=4,
s=7,
rasterized=True,
linewidths=0,
label='Core',
marker='.')
axs[1].plot(get_xval(target_df),
get_yval(target_df),
alpha=1,
mew=0.5,
zorder=8,
label='TOI 1937',
markerfacecolor='lightskyblue',
markersize=14,
marker='*',
color='black',
lw=0)
axs[1].set_ylim(axs[1].get_ylim()[::-1])
axs[1].set_xlabel('Bp - Rp [mag]')
axs[1].set_ylabel('Absolute G [mag]', labelpad=-6)
##########
words = ['Field', 'Corona', 'Core', 'TOI1937'][::-1]
colors = ['gray', 'lightskyblue', 'k', 'lightskyblue'][::-1]
rainbow_text(0.98, 0.02, words, colors, size='medium', ax=axs[0])
f.tight_layout(w_pad=2)
outpath = os.path.join(outdir, 'small_ngc2516.png')
savefig(f, outpath)
import os
from glob import glob
import pandas as pd, numpy as np
from astropy.table import Table
OVERWRITE = 0
CDIPSVER = 0.6
# Bedell's 1 arcsecond DR2 X Kepler match
KEPGAIAFUN = '/Users/luke/local/kepler_gws/kepler_dr2_1arcsec.fits'
df_kep = Table.read(KEPGAIAFUN, format='fits').to_pandas()
df_kep.source_id = df_kep.source_id.astype(str)
df_kep['planet?'] = df_kep['planet?'].str.decode('utf-8')
df_cdips = get_cdips_catalog(ver=CDIPSVER)
df_cdips['source_id'] = df_cdips.source_id.astype(str)
mdf = df_kep.merge(df_cdips, how='inner', on='source_id')
outdf = mdf[~pd.isnull(mdf.reference_id)]
outpath = os.path.join(DATADIR, '20210916_kepgaiafun_X_cdipsv0pt6.csv')
outdf.to_csv(outpath, index=False)
print(f'Wrote {outpath}')
selcols = [
'kepid', 'source_id', 'kepmag', 'phot_g_mean_mag_x', 'cluster',
'reference_id', 'mean_age', 'planet?', 'nkoi', 'nconfp',
'kepler_gaia_ang_dist'
]
/nfs/phtess2/ar0/TESS/PROJ/lbouma/CDIPS_LCS/get_lc_list.sh
"""
import os
import pandas as pd
from cdips.utils.catalogs import get_cdips_catalog
from cdips.utils import today_YYYYMMDD
lclistpath = f'/nfs/phtess2/ar0/TESS/PROJ/lbouma/CDIPS_LCS/lc_list_{today_YYYYMMDD()}.txt'
with open(lclistpath, 'r') as f:
lines = f.readlines()
lines = [l.rstrip('\n') for l in lines]
df = get_cdips_catalog()
sourceids = [l.split('/')[-1].split('gaiatwo')[-1].split('-')[0].lstrip('0') for l in lines]
lcdf = pd.DataFrame({'path':lines, 'source_id':sourceids})
lcdf.source_id = lcdf.source_id.astype(str)
df.source_id = df.source_id.astype(str)
mdf = lcdf.merge(df, on='source_id', how='left')
outdir = '/nfs/phtess2/ar0/TESS/PROJ/lbouma/cdips/results/lc_metadata'
outpath = os.path.join(
outdir, f'source_ids_to_cluster_merge_{today_YYYYMMDD()}.csv'
)
mdf.to_csv(outpath, index=False, sep=';')
def get_cluster_lc_paths(N_lcs=100):
"""
Draw light curves from a set of clusters that I think might be good in
S14-S19 to test planet injection/recovery.
These are in a list of 8 clusters, and have RP<14.
"""
from cdips.utils.catalogs import get_cdips_catalog
df = get_cdips_catalog(ver=0.6)
df = df[~pd.isnull(df.cluster)]
cluster_names = [
"Stephenson_1", # logt 7.5, 100-500pc, Theia 73, we are familiar
"Alpha_per", # logt 7.8, 100-200pc, Theia 133
"RSG_5", # logt 7.5, 300-400pc.
"kc19group_506", # 350 Myr, 100pc (!!!).
"kc19group_507", # 350 Myr, 100-200pc.
"AB_Dor", # ~120 Myr, 100-300pc from KC19.
"Pleiades", # 120 Myr, 120 pc.
]
special_names = [
"UBC_1", # 350 Myr, 300-400pc, Theia 520
]
sel = np.zeros(len(df)).astype(bool)
for cluster_name in cluster_names:
sel |= df.cluster.str.contains(cluster_name)
# avoid e.g., "UBC_186" matching for "UBC_1".
for cluster_name in special_names:
sel |= (df.cluster == cluster_name)
sel |= df.cluster.str.contains(cluster_name + ',')
sdf = df[sel]
# NOTE: a further cut, to deal with less crappy data
sel = sdf.phot_rp_mean_mag < 14
sdf = sdf[sel]
outdir = resultsdirectory
camdpath = os.path.join(outdir, f'camd_{"_".join(cluster_names)}.png')
if not os.path.exists(camdpath):
plt.close('all')
get_yval = (lambda _df: np.array(_df['phot_g_mean_mag'] + 5 * np.log10(
_df['parallax'] / 1e3) + 5))
get_xval = (lambda _df: np.array(_df['phot_bp_mean_mag'] - _df[
'phot_rp_mean_mag']))
fig, axs = plt.subplots(ncols=2)
s = 0.5
axs[0].scatter(get_xval(sdf),
get_yval(sdf),
c='k',
alpha=1,
zorder=3,
s=s,
rasterized=False,
linewidths=0.1,
marker='o',
edgecolors='k')
axs[0].set_xlabel("BP-RP [mag]")
axs[0].set_ylabel("M_G [mag]")
axs[1].scatter(get_xval(sdf),
sdf.phot_g_mean_mag,
c='k',
alpha=1,
zorder=3,
s=s,
rasterized=False,
linewidths=0.1,
marker='o',
edgecolors='k')
axs[1].set_xlabel("BP-RP [mag]")
axs[1].set_ylabel("G [mag]")
fig.savefig(camdpath, bbox_inches='tight', dpi=400)
# NOTE: OK! Now, find the light curves, from S14-S19, corresponding to
# these clusters.
from cdips.utils.lcutils import make_lc_list
lclistpath = (
"/nfs/phtess2/ar0/TESS/PROJ/lbouma/CDIPS_LCS/S14_S19_lc_list_20211019.txt"
)
if not os.path.exists(lclistpath):
make_lc_list(lclistpath, sector_interval=[14, 19])
lcdf = pd.read_csv(lclistpath, names=["lcpath"])
lcdf['source_id'] = lcdf.lcpath.apply(lambda x: x.split('/')[-1].split(
'gaiatwo')[1].split('-')[0].lstrip('0'))
sdf['source_id'] = sdf.source_id.astype(str)
mdf = sdf.merge(lcdf, on="source_id", how="left")
smdf = mdf[~pd.isnull(mdf.lcpath)]
res = Counter(smdf.cluster)
print(42 * '-')
print(
f'Got {len(smdf)} T<14 light curves in {cluster_names} & {special_names}...'
)
print(f'Most common 40:\n{res.most_common(n=40)}')
outpath = f'S14_S19_20211020_lcpaths_{"_".join(cluster_names)}.csv'
smdf.to_csv(outpath, index=False)
print(f'Saved {outpath}')
lcpaths = nparr(smdf.lcpath)
np.random.shuffle(lcpaths)
lcpaths = np.random.choice(lcpaths, size=N_lcs)
return lcpaths
def plot_TIC268_nbhd_full(outdir=RESULTSDIR):
set_style()
df = get_cdips_catalog(ver=0.4)
kc19_sel = ((df.cluster.str.contains('NGC_2516')) &
(df.reference.str.contains('Kounkel_2019')))
cg18_sel = ((df.cluster.str.contains('NGC_2516')) &
(df.reference.str.contains('CantatGaudin_2018')))
target_df = df[df.source_id == 5489726768531119616] # TIC 2683...
kc19_df = df[kc19_sel]
cg18_df = df[cg18_sel]
##########
plt.close('all')
params = ['ra', 'dec', 'parallax', 'pmra', 'pmdec']
nparams = len(params)
qlimd = {
'ra': 0,
'dec': 0,
'parallax': 0,
'pmra': 0,
'pmdec': 0,
'radial_velocity': 0
} # whether to limit axis by quantile
ldict = {
'ra': r'$\alpha$ [deg]',
'dec': r'$\delta$ [deg]',
'parallax': r'$\pi$ [mas]',
'pmra': r'$\mu_{{\alpha}} \cos\delta$ [mas/yr]',
'pmdec': r'$\mu_{{\delta}}$ [mas/yr]',
'radial_velocity': 'RV [km/s]'
}
f, axs = plt.subplots(figsize=(6, 6), nrows=nparams - 1, ncols=nparams - 1)
for i in range(nparams):
for j in range(nparams):
print(i, j)
if j == nparams - 1 or i == nparams - 1:
continue
if j > i:
axs[i, j].set_axis_off()
continue
xv = params[j]
yv = params[i + 1]
print(i, j, xv, yv)
axs[i, j].scatter(kc19_df[xv],
kc19_df[yv],
c='gray',
alpha=0.9,
zorder=2,
s=5,
rasterized=True,
linewidths=0,
label='Core',
marker='.')
axs[i, j].scatter(cg18_df[xv],
cg18_df[yv],
c='k',
alpha=0.9,
zorder=3,
s=5,
rasterized=True,
linewidths=0,
label='Corona')
axs[i, j].plot(target_df[xv],
target_df[yv],
alpha=1,
mew=0.5,
zorder=8,
label='TOI 1937',
markerfacecolor='yellow',
markersize=14,
marker='*',
color='black',
lw=0)
# set the axis limits as needed
if qlimd[xv]:
xlim = (np.nanpercentile(kc19_df[xv], 25),
np.nanpercentile(kc19_df[xv], 75))
axs[i, j].set_xlim(xlim)
if qlimd[yv]:
ylim = (np.nanpercentile(kc19_df[yv], 25),
np.nanpercentile(kc19_df[yv], 75))
axs[i, j].set_ylim(ylim)
# fix labels
if j == 0:
axs[i, j].set_ylabel(ldict[yv])
if not i == nparams - 2:
# hide xtick labels
labels = [
item.get_text()
for item in axs[i, j].get_xticklabels()
]
empty_string_labels = [''] * len(labels)
axs[i, j].set_xticklabels(empty_string_labels)
if i == nparams - 2:
axs[i, j].set_xlabel(ldict[xv])
if not j == 0:
# hide ytick labels
labels = [
item.get_text()
for item in axs[i, j].get_yticklabels()
]
empty_string_labels = [''] * len(labels)
axs[i, j].set_yticklabels(empty_string_labels)
if (not (j == 0)) and (not (i == nparams - 2)):
# hide ytick labels
labels = [
item.get_text() for item in axs[i, j].get_yticklabels()
]
empty_string_labels = [''] * len(labels)
axs[i, j].set_yticklabels(empty_string_labels)
# hide xtick labels
labels = [
item.get_text() for item in axs[i, j].get_xticklabels()
]
empty_string_labels = [''] * len(labels)
axs[i, j].set_xticklabels(empty_string_labels)
# axs[2,2].legend(loc='best', handletextpad=0.1, fontsize='medium', framealpha=0.7)
# leg = axs[2,2].legend(bbox_to_anchor=(0.8,0.8), loc="upper right",
# handletextpad=0.1, fontsize='medium',
# bbox_transform=f.transFigure)
for ax in axs.flatten():
format_ax(ax)
f.tight_layout(h_pad=0.1, w_pad=0.1)
outpath = os.path.join(outdir, 'full_kinematics.png')
savefig(f, outpath)