Skip to first unread message
danny...@g.harvard.edu
Oct 22, 2018, 5:13:49 PM10/22/18
to Dataverse Users Community
Hi everyone,
We'll have our regularly scheduled call tomorrow at Noon ET. For more details, including call in information, check out the usual page:
See you soon!
- Danny
Philip Durbin
Oct 24, 2018, 4:20:49 PM10/24/18
to dataverse...@googlegroups.com
Great meeting yesterday! Here are the notes from https://docs.google.com/document/d/1wksJrrp-vU2SRLkqqBFzx5_Dc6bGMhqh-l0L8H5VMUw/edit?usp=sharing
2018-10-23 Dataverse Community Call
Agenda
* Community Questions
* Reproducibility Tools (https://global.gotomeeting.com/join/899867157)
Attendees
* Danny Brooke (IQSS)
* Philip Durbin (IQSS)
* Tania Schlatter (IQSS)
* Matthew Dunlap (IQSS)
* Kevin Condon (IQSS)
* Travis Hewgley (Code Ocean)
* Shahar Zaks (Code Ocean)
* Russ (Code Ocean)
* Jim Myers (QDR, TDL)
* Pete Meyer (HMS)
* Courtney Mumma (TDL)
* Slava (DANS)
* Amber Leahey (SP)
* Kaitlin Newson (SP)
* Sherry Lake (UVa)
* James Sappenfield (IQSS / HBS)
* James Turrito (JPAL)
Notes
* Community Questions
* (Courtney) Uptick in spiders/disruptive crawlers? 4.8.6
* Increase in the last week, community post incoming. Update: posted at https://groups.google.com/d/msg/dataverse-community/qCUw2uZ1feE/PoxFhkwqCAAJ
* (Pete) Noted a crawler that does not respect robots.txt (BLEXBot; blocked it)
* (Pete) did not see problems with system stability
* Crawlers going after schema.org - pipl, vultr, semrush ...
* Sloan Grant Reproducibility Tools
* (Danny) Background on the Sloan grant: https://groups.google.com/d/msg/dataverse-community/k4x6lhaxGpc/bK6V2THzFgAJ
* (Travis) Showing https://codeocean.com/2018/10/16/the-contact-hypothesis-re-evaluated-colon-code-and-data/code and "compute capsule" embedded at https://www.cambridge.org/core/journals/behavioural-public-policy/article/contact-hypothesis-reevaluated/142C913E7FA9E121277B29E994124EC5
* (Pete) Is there a way to specify compute requirements?
* (?) Default machine has good specs.
* Not yet; more info in knowledge base
* (?) Question about user interface and how Code Ocean and Dataverse would interact. How would it look? If you had your data on Dataverse, would Code Ocean accept the DOI from Dataverse? Would it work like TwoRavens?
* (Danny) Initial thought is to use the TwoRavens model as an "explore" tool.
* (Shahar) Code Ocean would work when there is both code and data. You might need to make some adjustments to your code and data to make it executable on Code Ocean. Focus on reproducibility. Code Ocean only works on data coupled with code.
* (Pete) Different academic research groups collaborating. Sometimes they can't share data. What if the executables are not part of the dataset?
* (Shahar) Code Ocean only allows you to publish code that executes. Demonstrating usage.
* Access Stata, MatLab
* Potential for arrangement between code ocean and academic dev groups to report code usage (for dev funding agencies)
* (Travis) Creating a compute capsule data using Python, showing a line graph. Install R. Both in the capsule. Support for RStudio, Jupyter notebooks. Can select and apply a license. Software license is separate from Data license. Can fetch publication from a DOI. Can invite collaborators.
* (Jim) Standards are emerging for downloading data from a DOI.
* (Phil) There is a conversation of downloading files based on a dataset DOI using Schema.org JSON-LD fields at https://github.com/whole-tale/whole-tale/issues/35#issuecomment-427411937
* (Slava) Long term strategy for preservation?
* (Shahar) Wrapping up a preservation document.
* Working with CLOCKSS
* Export feature available from Code Ocean. Docker file.
* (Danny) Can we see the "import code from GitHub" button?
* (Shahar) Works with any public repository.
* (Slava) Is it possible to export the complete running container?
* (Shahar) Yes. Can include the data or not.
* (Travis) Metadata comes out of the YAML file.
* (James) Study with Andre on reproducibility, progress of science. Any idea on what implementation would look like? Any plan for a study? Sounds like the plan is roll out the implementation on a voluntary basis.
* (Danny) Room in the grant questions would be answered by Merce.
* (Danny) In Dataverse we track explorations per tool.
* (Danny) Each installation of Dataverse would decide if they want to set up an integration with Code Ocean.
* (Danny) How does the Dataverse community keep in touch?
* (Travis) Please feel free to contact me: tra...@codeocean.com
Agenda
* Community Questions
* Reproducibility Tools (https://global.gotomeeting.com/join/899867157)
Attendees
* Danny Brooke (IQSS)
* Philip Durbin (IQSS)
* Tania Schlatter (IQSS)
* Matthew Dunlap (IQSS)
* Kevin Condon (IQSS)
* Travis Hewgley (Code Ocean)
* Shahar Zaks (Code Ocean)
* Russ (Code Ocean)
* Jim Myers (QDR, TDL)
* Pete Meyer (HMS)
* Courtney Mumma (TDL)
* Slava (DANS)
* Amber Leahey (SP)
* Kaitlin Newson (SP)
* Sherry Lake (UVa)
* James Sappenfield (IQSS / HBS)
* James Turrito (JPAL)
Notes
* Community Questions
* (Courtney) Uptick in spiders/disruptive crawlers? 4.8.6
* Increase in the last week, community post incoming. Update: posted at https://groups.google.com/d/msg/dataverse-community/qCUw2uZ1feE/PoxFhkwqCAAJ
* (Pete) Noted a crawler that does not respect robots.txt (BLEXBot; blocked it)
* (Pete) did not see problems with system stability
* Crawlers going after schema.org - pipl, vultr, semrush ...
* Sloan Grant Reproducibility Tools
* (Danny) Background on the Sloan grant: https://groups.google.com/d/msg/dataverse-community/k4x6lhaxGpc/bK6V2THzFgAJ
* (Travis) Showing https://codeocean.com/2018/10/16/the-contact-hypothesis-re-evaluated-colon-code-and-data/code and "compute capsule" embedded at https://www.cambridge.org/core/journals/behavioural-public-policy/article/contact-hypothesis-reevaluated/142C913E7FA9E121277B29E994124EC5
* (Pete) Is there a way to specify compute requirements?
* (?) Default machine has good specs.
* Not yet; more info in knowledge base
* (?) Question about user interface and how Code Ocean and Dataverse would interact. How would it look? If you had your data on Dataverse, would Code Ocean accept the DOI from Dataverse? Would it work like TwoRavens?
* (Danny) Initial thought is to use the TwoRavens model as an "explore" tool.
* (Shahar) Code Ocean would work when there is both code and data. You might need to make some adjustments to your code and data to make it executable on Code Ocean. Focus on reproducibility. Code Ocean only works on data coupled with code.
* (Pete) Different academic research groups collaborating. Sometimes they can't share data. What if the executables are not part of the dataset?
* (Shahar) Code Ocean only allows you to publish code that executes. Demonstrating usage.
* Access Stata, MatLab
* Potential for arrangement between code ocean and academic dev groups to report code usage (for dev funding agencies)
* (Travis) Creating a compute capsule data using Python, showing a line graph. Install R. Both in the capsule. Support for RStudio, Jupyter notebooks. Can select and apply a license. Software license is separate from Data license. Can fetch publication from a DOI. Can invite collaborators.
* (Jim) Standards are emerging for downloading data from a DOI.
* (Phil) There is a conversation of downloading files based on a dataset DOI using Schema.org JSON-LD fields at https://github.com/whole-tale/whole-tale/issues/35#issuecomment-427411937
* (Slava) Long term strategy for preservation?
* (Shahar) Wrapping up a preservation document.
* Working with CLOCKSS
* Export feature available from Code Ocean. Docker file.
* (Danny) Can we see the "import code from GitHub" button?
* (Shahar) Works with any public repository.
* (Slava) Is it possible to export the complete running container?
* (Shahar) Yes. Can include the data or not.
* (Travis) Metadata comes out of the YAML file.
* (James) Study with Andre on reproducibility, progress of science. Any idea on what implementation would look like? Any plan for a study? Sounds like the plan is roll out the implementation on a voluntary basis.
* (Danny) Room in the grant questions would be answered by Merce.
* (Danny) In Dataverse we track explorations per tool.
* (Danny) Each installation of Dataverse would decide if they want to set up an integration with Code Ocean.
* (Danny) How does the Dataverse community keep in touch?
* (Travis) Please feel free to contact me: tra...@codeocean.com
--
You received this message because you are subscribed to the Google Groups "Dataverse Users Community" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dataverse-commu...@googlegroups.com.
To post to this group, send email to dataverse...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dataverse-community/ac3e88cd-6f01-4b55-8b9c-bbb0db78d081%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
--
Philip Durbin
Software Developer for http://dataverse.org
http://www.iq.harvard.edu/people/philip-durbin
Software Developer for http://dataverse.org
http://www.iq.harvard.edu/people/philip-durbin
Reply all
Reply to author
Forward
0 new messages