HI!

THis is the social metrics project where we are iterating on creating graphs to answer questions about social support.

Draft #highest number is the most recent creating# #index has all the questions that we want to answer in the program with the current export data that we can access and report to others at Mozilla#

#All of the data is private, but is public information on Twitter#

Latest update Draft 6 folder: Week 1 62 Draft count charts included in report

1. Input -p -t -c -f required
2. Trending tags are pulled from the load of the first chart
3. This is data from the first week of a release <2500 for the datatype (furture larger data sets may be problematic)
4. Action Items:

• comparing week to week = new chart needed
• explore jupiter notebooks and hosting for reports
• explore word count and positive or negative support experiences

Latest update Draft 5 folder: Needed: (instructions on how to prepare the csv file)

1. Whole directory
2. Updated tending tag file the txt should include a new tag created on each line
3. Export conversations details from Reports section of Reply by Buffer (one day or week of data)
4. Filter the css output to just Help Me and AoA tags (this is what we identified as ‘support conversations’)
5. Run the script with this new css file and with all of the tag files
6. Save the graphs 7.Save Data description
7. Read all conversations and add notes to describe each category
8. If solved - a reply time will be present (search in spreadsheet for tag for the fastest way to find )

To run this script with the daily output use the command below:

MacBook-Pro-20:Draft5 rmcguigan$ python tagcharts.py -f Week1Sept5.csv -s supporttags.txt -p producttags.txt -t trendingtags.txt -l languagetags.txt

Three graphs will generate, save each one then close them.

This will appear in the console - record this for the data description section of your report totals ^

{'Help Me': 44, 'AoA': 11} {'New Release': 2}

Good Luck!

Draft 8 This version compares 3 different dumps for 3 weeks to compare Categories, Language and Product volume. It runs with jupyter notebooks and requires it to be running for the nyp file to open in a browser window. The charts are created outside of the notebook. Bugs: need import library to show charts in notebook, need hosting for csv dumps for daily keyword monitoring.

Open Lisence