A Portable Data Analysis Package for MUSTANG Metrics in the Field
With MUSTANG having reached a certain level of production maturity and usefulness to the end user, we have been receiving more requests for access to the source code that generates the data quality metrics, as well as some means of easily driving these metrics in a scripted fashion.
IRIS Data Services began pondering a project to create a MUSTANG-lite that users could download and use on their own data, or even from data provided by FDSN web services. This project took on the name of ISPAQ, which is pronounced like Ice Pack, sure to relieve the headaches of many scientists and principal investigators that are looking for a simple package to run quality metrics on their own data.
Making use of Python, obspy, and R, ISPAQ is a portable client package that will offer lots of versatility and will be easy to script and automate. Making use of user-defined preferences for metrics, stations, and source data, ISPAQ is easily run from the command line. The following example shows ISPAQ being run multiple times in series to generate CSV files.
2016-07-14 08:57:14 – INFO – ALL FINISHED!
2016-07-14 08:57:17 – INFO – Running ISPAQ version 0.7.2 on Thu Jul 14 08:57:17 2016
2016-07-14 08:57:17 – DEBUG – Creating UserRequest …
2016-07-14 08:57:17 – DEBUG – Validating preferred metrics …
2016-07-14 08:57:17 – DEBUG – Creating Concierge …
2016-07-14 08:57:17 – DEBUG – Inside simple business logic …
2016-07-14 08:57:17 – INFO – Calculating simple metrics for 1 SNCLs.
2016-07-14 08:57:17 – INFO – 000 Calculating simple metrics for IU.ANMO.00.BHZ
2016-07-14 08:57:25 – INFO – Writing simple metrics to test_simple_IU.ANMO.00.BHZ_2016-04-20__simpleMetrics.csv.
2016-07-14 08:57:25 – INFO – ALL FINISHED!
2016-07-14 08:57:27 – INFO – Running ISPAQ version 0.7.2 on Thu Jul 14 08:57:27 2016
2016-07-14 08:57:27 – DEBUG – Creating UserRequest …
2016-07-14 08:57:27 – DEBUG – Validating preferred metrics …
2016-07-14 08:57:27 – DEBUG – Creating Concierge …
2016-07-14 08:57:27 – DEBUG – Inside simple business logic …
2016-07-14 08:57:27 – INFO – Calculating simple metrics for 9 SNCLs.
2016-07-14 08:57:27 – INFO – 000 Calculating simple metrics for II.EFI.00.BH1
2016-07-14 08:57:35 – INFO – 001 Calculating simple metrics for II.EFI.00.BH2
2016-07-14 08:57:42 – INFO – 002 Calculating simple metrics for II.EFI.00.BHZ
2016-07-14 08:57:49 – INFO – 003 Calculating simple metrics for II.ERM.00.BH1
2016-07-14 08:57:56 – INFO – 004 Calculating simple metrics for II.ERM.00.BH2
2016-07-14 08:58:04 – INFO – 005 Calculating simple metrics for II.ERM.00.BHZ
2016-07-14 08:58:11 – INFO – 006 Calculating simple metrics for II.ESK.00.BH1
2016-07-14 08:58:17 – INFO – 007 Calculating simple metrics for II.ESK.00.BH2
2016-07-14 08:58:24 – INFO – 008 Calculating simple metrics for II.ESK.00.BHZ
2016-07-14 08:58:31 – INFO – Writing simple metrics to test_simple_crossCorrelation_2016-04-20__simpleMetrics.csv.
2016-07-14 08:58:31 – INFO – ALL FINISHED!
2016-07-14 08:58:33 – INFO – Running ISPAQ version 0.7.2 on Thu Jul 14 08:58:33 2016
2016-07-14 08:58:33 – DEBUG – Creating UserRequest …
2016-07-14 08:58:33 – DEBUG – Validating preferred metrics …
2016-07-14 08:58:33 – DEBUG – Creating Concierge …
2016-07-14 08:58:33 – DEBUG – Inside simple business logic …
2016-07-14 08:58:33 – INFO – Calculating simple metrics for 1 SNCLs.
2016-07-14 08:58:33 – INFO – 000 Calculating simple metrics for US.OXF.00.BHZ
2016-07-14 08:58:49 – INFO – Writing simple metrics to test_simple_test_small_quake_w_gaps_2016-04-20__simpleMetrics.csv.
The resulting output is easy to make use of within R, Matlab, and Excel to carry out tabulation, computation, and plotting, all taking place on your own client machine, using your own scripted workflows. ISPAQ generates its own metrics values from the raw data, independent of MUSTANG web services.
ISPAQ has been in production for the past few months, with parallel improvements made to the R code to support MUSTANG and ISPAQ jointly. We are looking at an autumn release of ISPAQ to the public, with the code being fully open-source, in tandem with an updated release to the R code on CRAN for the IRISMustangMetrics, IRISSeismic, and seismicRoll packages.
ISPAQ is designed to make use of the latest version of the IRIS R packages on CRAN and will be capable of detecting and adapting to metrics code updates as they are deployed. This will help ensure that the user in the field has the latest and complete version of the metrics code that is also in production for MUSTANG.
Look for future announcements on the upcoming ISPAQ release by going to the IRIS Message Center and signing up for the DMC Software mailing list.
http://ds.iris.edu/message-center/
by Rob Casey , Tim Ahern and Gillian Sharer (IRIS Data Management Center)