On this week I was make the following issues:
- Fix repository scripts: This repository contains the scripts to helper the work with AppRecommender and the popularity-contest data, one of these scripts, its the script to generate clusters of popularity-contest data. link to repository:
https://gitlab.com/TCC-AppRecommender/scripts - Change save popcon clusters method: Before the data generated by the script was saved with python-pickle, now this data is saved with text on files, to be used with any language
- Update AppRecommender load popcon clusters: The AppRecommender was updated to read the files without pickle, but with text on files, with this the tests have also been updated.
- Optimize memory consumption on collect popcon data: Before the script was consuming more than 7% of the memory for read 1300 submissions.
Now the script is consuming better than 1.4% of the memory for read 1300 submissions. The memory of my PC is 8GB
For the next week the focus is continue improving the memory consumption, and if possible improve the time for read the popularity-contest submissions and research about the security of users data