Author Archives: thomas

MSNoise 1.24 released

The latest release of MSNoise has been published, it includes:

Thanks to some great early MSNoise adopters who reported problems using the mailing-list, we have identified a few bugs and tricky situations where some steps failed. This is the case for

    scan_archive: there were major issues using Threads, this step uses Process for multiprocessing, which is much safer. The only problem remaining is that there is no more console-logging of the found/identified files. To be corrected soon.

    new_jobs: mostly rewritten, all jobs are properly identified now. There should no more lost jobs !

I've also reflected those changes in the documentation.

As usual, it is available via GitHub:

MSNoise 1.2

I’m thrilled to announce the release of MSNoise 1.2 ! This release marks a big change in many aspects! In terms of performances:

  • The MySQL DB is no longer a “forced choice”: one can use a simple sqlite database to run MSNoise. We still strongly recommend using MySQL for performance reasons.
  • The whiten function now returns the FFT of the signal, as this is directly used by MyCorr, and we save the time needed to ifft then fft the array.
  • The database_tools have been completely rewritten in order to use SQLAlchemy instead of explicit MySQL expressions.

Thanks to the feedback of some early users, we have been able to track down bugs and pitfalls along the steps. The documentation is now linked directly to the code, mostly embedded in the steps, which allows its evolution together with the code. I have rewritten the Installation Guide, it’s now quite complete for Windows and Linux. I need to find a Mac to test it sometime soon. There are still some steps of the installation procedure that are not straightforward for most users (MySQL server set-up), I’ll try to provide guidelines for this very soon.

For those already addicted to MSNoise, note the steps have changed names: for example 02.new_jobs.py is now s02new_jobs.py. This allows importing each step as a module, which was required for building the documentation.

MSNoise 1.2 is available on GitHub. Download it and start playing ! Note, 1.2 is not compatible with the previous version (well, it *should* work with a little tweaking to the database, scream on the Mailing-List if you need help!).

Thanks to Kasper, Zack, Carolin, Philippe and Agathe for their inputs!

Install video (using sqlvolution branch)

I’ve produced a small video of MSNoise, from the download of the current ‘sqlvolution’ branch until the production of 1 single CCF. Be sure to select the 720p quality !

The video shows the following steps:

  1. download of the zip file
  2. installation (=creation of the database “msnoise.sqlite”)
  3. configuration
  4. scanning of the archive
  5. definition of new jobs
  6. usage of the SQLite Manager extension for Firefox to check the content of the database/tables
  7. run of the computation of the CCF
  8. and finally a plot of the CCF using ipython+obspy

and all this in … 4 minutes !

MSNoise package available

As promised, the MSNoise package is available for download (see the Installation page). If you’re using it, please send us an email, so we can keep you updated of the developments. We will soon set up a mailing list, probably using GoogleGroups, so keep in touch !

Disclaimer

Although we have cross-checked the whole code, we cannot warranty it is exempt of bugs. The package is provided as-is, we will not be held responsible for any use you make of it, nor for the results and conclusions you may find using MSNoise.

Licence

Although MSNoise is provided freely on the internet and every individual may use it, we do not allow anyone to provide commercial service (read -paid in any way-), like training or teaching without contacting the original authors first. The authors and the contact email address are in the __init__.py file of the package.