An unusual experiment in collaborative reproducible research took place in Austin, Texas, on July 25-27: 25 participants from 9 different organizations gathered at the Bureau of Economic Geology, The University of Texas at Austin for the First Madagascar Working Workshop. The participants divided into 10 teams of 2-3 people by pairing experienced Madagascar developers with novice users. Each team worked on a small project, updating either reproducible papers or entries in the migration gallery. At the conclusion of the meeting, the participants discussed their experience and plans for future workshops.
Most of the time at the workshop was spent in workgroups of two or three. Each workgroup has it's own story to tell. Karl, Medhi, and Xuan formed a group to migrate a 3D field dataset, TeaPot dome. Here is a summary of our three days of activities.
Day 1- Karl and Medhi met and started our project. We read documentation and learned sfzomig3 is post stack 3D migration based on the extended split step method. We found a prestack dataset with preprocessing to stack to use for input data. We also found a dmo stack velocity text file. In the afternoon we stacked the data using a single velocity function.
Day 2- Xuan joined our group and we spent a little time bringing him up to speed. We converted the velocity from RMS to interval and started the migration. We left the program running at the end of the day.
Day 3- The migration from day 2 was difficult to evaluate. The input was in time, while the output was in depth. We also bandlimitted the migration to 20 hz. We vertically stretched the migrated data to time, and bandpass filterred the stack. The migration looked correct. Most of the rest of the day we made sure the scripts were in the repository.
We plan to check the figures into the repository. Diffraction imaging would be a good program to try next.
1. 100% of those who responded to the survey would be interested in attended a Madagascar event in the future and would recommend it to their colleagues.
2. Some of the common things they liked about the workshop: (a) team work, pair programming ("Pair programming. First time for me.", "Collaboration spirit; Meet with other fantastic developers and users; Assistance and no-disturbing flag;") (b) the idea of migration gallery ("I think it is a good idea to compare different migration methods with different benchmark models.") (c) Food ("Good dinner.")
3. Some of the common things they didn't like about the workshop: (a) lack of preparation ("Many of us were not well prepared for the workshop.", "We could have used email to identify projects before the workshop started.", "I believe if we can do more prepared work before workshop, it may be more efficient. For example, the team leader, who want to start a project, can fist give a short introduction about his/her project. So that, he/she can recruit his/her team members.") (b) lack of group discussion ("Need more acitive discussion.") (c) lack of wifi ("We should have some wireless in the room.")
4. The most popular topic for a future working workshop: Parallel Computing (80% of respondents)
5. The most popular location for a future working workshop: Houston (65% of respondents)