|
Size: 859
Comment:
|
Size: 1956
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 9: | Line 9: |
| * You guys, Amazon '''finally''' added [[https://boto3.readthedocs.io/en/latest/reference/services/mturk.html|MTurk support to Boto3]] ([[https://blog.mturk.com/developers-can-now-use-aws-sdks-to-access-the-mturk-api-13eb557af2cc#.h0gx6buwf|blog post]]). If you want current MTurk API, you have to switch to boto3, but it's going to require some code rewriting due to changes between boto and boto3. --alw | * You guys, Amazon '''finally''' added [[https://boto3.readthedocs.io/en/latest/reference/services/mturk.html|MTurk support to Boto3]] ([[https://blog.mturk.com/developers-can-now-use-aws-sdks-to-access-the-mturk-api-13eb557af2cc#.h0gx6buwf|blog post]]). If you want current MTurk API, you have to switch to boto3, but it's going to require some code rewriting due to significant changes between boto and boto3. --alw * [[http://www.sciencemag.org/news/2016/06/psychologists-grow-increasingly-dependent-online-research-subjects|Science news post]] about MTurk pool and psych researchers' dependency on MTurk --alw |
| Line 18: | Line 19: |
| 1. Read the documentation for and played around with boto3, including trying to rewrite my getResults.py script. On the plus side, boto3 finally integrates MTurk with the rest of AWS, On the minus side, what you get back is no longer Python types wrapping MTurk XML data structures, but rather Python dictionaries of the keys and values parsed from the XML data. The `Answer` part of an `Assignment` actually still comes back as a string of `QuestionFormAnswers` XML you have to parse yourself. (Hopefully that changes down the road; MTurk support is only 2 weeks old in boto3.) 1. Started writing a proper README for [[https://bitbucket.org/hlplab/crowdexp|CrowdExp]] 1. Did a code cleanup of [[https://bitbucket.org/hlplab/mturkutils|MTurk Utils]] code 1. Started to really dig into the results conversion code in Crowd``Exp so it can dump R analyzable results files soon |
Lab Meeting, Spring 2017, Week 10
Awe- and Aw-inspiring readings (or just stuff you think deserves a mention)
You guys, Amazon finally added MTurk support to Boto3 (blog post). If you want current MTurk API, you have to switch to boto3, but it's going to require some code rewriting due to significant changes between boto and boto3. --alw
Science news post about MTurk pool and psych researchers' dependency on MTurk --alw
What we did over last week
Florian
Xin
Andrew
Read the documentation for and played around with boto3, including trying to rewrite my getResults.py script. On the plus side, boto3 finally integrates MTurk with the rest of AWS, On the minus side, what you get back is no longer Python types wrapping MTurk XML data structures, but rather Python dictionaries of the keys and values parsed from the XML data. The Answer part of an Assignment actually still comes back as a string of QuestionFormAnswers XML you have to parse yourself. (Hopefully that changes down the road; MTurk support is only 2 weeks old in boto3.)
Started writing a proper README for CrowdExp
Did a code cleanup of MTurk Utils code
Started to really dig into the results conversion code in CrowdExp so it can dump R analyzable results files soon
Jenn
Dan
Amanda
Zach
Linda
Wednesday