Use the generalized inversion technique to estimate stress drop for the 2019 Ridgecrest sequence. Original data is downloaded from Community Stress Drop Validation Study. gmprocess is used to download station files, event files, and then run QA on original data. Code modified from Klimasewski et al. (2019) is then used to run GIT.

Working directory henceforth denoted as ~/
QA processing takes place in ~/gmprocess
GIT takes place in ~/dataset_name

gmprocess:

        ~/gmprocess
                |
                +-- event_downloads
                |  	|
                |  	+-- conf
                |  	|	  |
                |  	|	  +-- config.yml
                |	+-- data
                |
                +-- qa_processing
                        |
                        +-- conf
                        |	  |
                        |	  +-- config.yml
                        +-- data

dataset:

        ~/dataset_name
                |
                +-- event_files
                |
                +-- station_data
                |
                +-- RC_beta
                |
                +-- RC_phase_beta
                        |
                        +-- eventid.phase

        /GitHub/stress_drop/
                |
                +-- set_up
                |	  |
                |	  +-- cp_event_files.py
                |	  |
                |	  +-- cp_eventjson.py
                |	  |
                |	  +-- cp_events.py
                |	  |
                |	  +-- cp_stn_files.py
                |  	  |
                |  	  +-- create_event_dirs.py
                |  	  |
                |	  +-- h5_to_mseed.py
                |
                +-- station_info
                |	  |
                |	  + stations.py
                |
                +-- event_info
                |	  |
                |	  + event_info.py
                |	
                +-- GIT
                      |
                      +-- step1_compute_spectra.py
                      |
                      +-- step2_secondo_meters.py
                      |
                      +-- step3_findBrune_trapezoids.py
                      |
                      +-- step4_secondo_constraint.py
                      |
                      +-- step5_fitBrune.py

1. Use create_event_dirs.py to create event directories for the dataset in both:
   a. ~/gmprocess/event_downloads/data/
   b. ~/dataset/RC_beta/
2. Use cp_events.py to copy original event data from its directory into the ~/gmprocess/qa_processing directory.
3. Download event.json files
   a. Enter ~/gmprocess/event_downloads directory and initialize gmprocess project.
   b. Set config.yml file downloader to very small radius in degrees (may have to do this more than once, weird stuff goes on with these *.yml files.
   c. Run >>gmrecords download.
   d. This will download event data for each event into its raw folder and an event.json file for each event. All we need from this download is that event.json file.
4. Use cp_eventjson.py to copy event.json files from event_downloads to their corresponding event directory in ~/gmprocess/qa_processing/data.
5. Use cp_stn_files.py to copy station *.xml files from ~/gmprocess/station_downloads/station_files into each event raw directory in ~/gmprocess/qa_processing/data.
6. Process dataset:
   a. Enter ~/gmprocess/qa_processing/ and initialize gmprocess.
   b. Set up config.yml file as desired for QA.
   c. Run >>gmrecords assemble.
   d. Run >>gmrecords process.
   e. Result is a *.h5 file in each event directory containing all waveforms and information about whether they passed our tests.
7. Use extract_tr.py to read *.h5 files and write all waveforms that passed QA screening to event directories in ~/dataset/processing/. NOTE: This python script must be run INSIDE an instance of gmprocess from command line!
8. Use cp_event_files.py to copy event *.txt files into ~/dataset/event_files.
9. Use stations.py to create *.csv file containing list of stations, locations, and how many times each station appears in the dataset. Records from stations that appear less than 3 times will be discarded in a later step. This file goes in ~/dataset/station_data.
10. Inversion:
    a. Use step1_comput_spectra.py to compute fft of data in ~/dataset/RC_beta/ and save them to ~/dataset/record_spectra/.
    b. Use step2_secondo_meters.py to obtain event and site spectra and save them to ~/dataset/Andrews_inversion/.
    c. Use step3_findBrune_trapezoids.py to find the most "Brune-like" (in shape) event spectrum to use as an amplitude constraint and save it to ~/dataset/constraint/.
    d. Use step4_secondo_constraint.py to apply constraint to all event and site spectra and save them to ~/dataset/Andrews_inversion_constrained/.
    e. Use step5_fitBrune.py to fit each event spectrum to the Brune model using nonlinear least squares, finding a best fit corner frequency and moment and saving the results in ~/dataset/stress_drops/stress_drops_dataset.csv.
11. Analyze results as desired.