CKS-Cool

1. Computing Environemnt
2. Target List Construction
3. Cookbook For CKS-X
4. Data Access
5. Isoclassify Cadence
6. Kepler Headers

Computing environment

Use the ckscool environment on Petigura's laptop

conda activate ckscool 
conda install numpy==1.15.4 # this avoids the ValueError: cannot set WRITEABLE flag to True of this array #24839
conda install scipy matplotlib astropy pandas d seaborn scikit-learn pytables
conda install -c conda-forge healpy # needed for dustmaps also got healpy==1.11 to work
conda install joblib # needed for occurrence valley work
pip install h5py # conda doesn't seem to work
pip install mwdust 
pip install pyephem
pip install lmfit
pip install ebfpy

isoclassify on f6f16e

Cookbook for target list construction

1. Construct target list
2. Plan/execute observations

Cookbook for occurrence analysis

• Compute stellar/planet parameters
• Access the CKS-Cool dataset

Assemble the data

1. Copy over CKS-Gaia HDF file

   cp ../CKS-Gaia/load_table_cache.hdf data/cksgaia_cache.hdf

2. Generate a symlink to dr25-chains_trimmed-thinned.hdf

   Note the DR25 chains are created in Kepler-Radius-Ratio repo. See its readme file. Users looking to rerun code, will need to get the datafile from Petigura.

3. Query Gaia parameters. Instructions here

Compute Stellar/Planetary Properties

1. Download smemp and smsyn catalogs to data/CKS_Spectroscopic_Parameters.csv

   https://jump.caltech.edu/explorer/58/
   mv ~/Downloads/CKS_Spectroscopic_Parameters.csv data/

2. Run the isoclassify code on combined DR1+DR2 dataset.

   Run the isoclassify code in three modes on three different spectroscopic parameters. On the combined DR1+DR2 data

   First generate the csv files of stellar parameters

   bin/run_ckscool.py create-iso-batch # creates 9 csv files
   

   Note, if you need to run just a few new stars, add line inrun_ckscool.py

   conda activate isoclassify

   Look at the logs and confirm isoclassify is behaving right. Run them on Erik's laptop.

   isoclassify multiproc direct 6 data/isoclassify-smsyn-direct.csv isoclassify/smsyn/direct.csv --baseoutdir isoclassify/smsyn/direct/  --plot none
   isoclassify multiproc grid 6 data/isoclassify-smsyn-grid-parallax-yes.csv isoclassify/smsyn/grid-parallax-yes.csv --baseoutdir isoclassify/smsyn/grid-parallax-yes/ --plot none
   isoclassify multiproc grid 6 data/isoclassify-smsyn-grid-parallax-no.csv isoclassify/smsyn/grid-parallax-no.csv --baseoutdir isoclassify/smsyn/grid-parallax-no/ --plot none
   isoclassify multiproc direct 6 data/isoclassify-smemp-direct.csv isoclassify/smemp/direct.csv --baseoutdir isoclassify/smemp/direct/ --plot none
   isoclassify multiproc grid 6 data/isoclassify-smemp-grid-parallax-yes.csv isoclassify/smemp/grid-parallax-yes.csv --baseoutdir isoclassify/smemp/grid-parallax-yes/ --plot none
   isoclassify multiproc grid 6 data/isoclassify-smemp-grid-parallax-no.csv isoclassify/smemp/grid-parallax-no.csv --baseoutdir isoclassify/smemp/grid-parallax-no/ --plot none

   conda deactivate

   Them create isoclassify tables

   run_ckscool.py create-iso-table

Generate ReaMatch table

Run ReaMatch.ipynb and copy output file to reamatch.csv to ~/Dropbox/CKS-Cool/hires/reamatch.csv. Howard Isaacson will then add the appropriate RM designations to file.

Generate representative spectra figure

Run the 3_Spectra-Figure ipython notebook

rsync -av --progress --files-from=data/fig_spectra/fig_spectra-files.txt cadence:/ data/fig_spectra/ 

Run the gap fitting code.

Takes about 1 hour to run with 1000 samples

bin/run_ckscool.py create-gradient # stores results in analysis/ to do a fresh run, remove or move this directory

Calling this will generate a bunch of other cached results

Create plots and build paper

run_ckscool.py build plot all run_ckscool.py build table all run_ckscool.py build val all run_ckscool.py build csv all

http://localhost:8889/notebooks/gapfitting-comparison/gapfitting.ipynb

Access to the CKS-Cool dataset

The full list of star and planet properties are in data/ckscool-planets-cuts.csv see data/column-definitions.txt for a description of the columns.

The is* columns correspond to cuts. See the ckscool/cuts.py for additional info.

Running isoclassify on cadence

Note when running on cadence, there was a really weird issue with h5py. Where it was taking 30s to read in the Combined Dustmap

Download all Kepler headers (this is needed for the dilution cut)

1. Get bactch download script here

2. Combine wget scripts and cat them into one file and only pull the fits files.

   cat ~/Downloads/Kepler_Quarterly_wget/Kepler_Q* | grep fits > Kepler_wget.bat

3. Download onto cadence:/data/user/petigura/lightcurve-all cadence Should take about 4 hours to download all.

   scrape_headers.py