Closed JamesKunstle closed 11 months ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
View / edit / reply to this conversation on ReviewNB
oindrillac commented on 2023-10-12T12:35:10Z ----------------------------------------------------------------
same comment here as the last PR on a brief interpretation of this graph and maybe you can clarify in the comments which color is which group?
JamesKunstle commented on 2023-11-07T18:28:49Z ----------------------------------------------------------------
I'll be rebasing this notebook only the final merged 'repo_discovery' notebook that has a longer conclusion at the end here, describing this figure and noting some ideas for next steps, including those reported below.
View / edit / reply to this conversation on ReviewNB
oindrillac commented on 2023-10-12T12:35:11Z ----------------------------------------------------------------
Can we link the document here for explanation of alpha and beta?
For a tldr of logic, a short explanation at the top here of alpha/beta would help. Maybe just the formula here and intuition behind it?
hemajv commented on 2023-10-13T18:42:46Z ----------------------------------------------------------------
++ I also think a brief explanation of what alpha/beta is and why we are defining it would be useful here
JamesKunstle commented on 2023-11-07T18:29:07Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
oindrillac commented on 2023-10-12T12:35:13Z ----------------------------------------------------------------
What do high and low alpha scores signify? How would you interpret the plot here? The alpha values look even for all projects? Any outliers that you observed?
JamesKunstle commented on 2023-11-07T19:09:20Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
oindrillac commented on 2023-10-12T12:35:14Z ----------------------------------------------------------------
This is very helpful! thanks ++
JamesKunstle commented on 2023-11-07T19:14:38Z ----------------------------------------------------------------
++
++ I also think a brief explanation of what alpha/beta is and why we are defining it would be useful here
View entire conversation on ReviewNB
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:05:57Z ----------------------------------------------------------------
Line #4. n_known = joined_counts["count_known"].sum()
so this is the "total" number of contribution events (summed over all the repositories) made by the sub-population?
yep!
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:05:58Z ----------------------------------------------------------------
could you maybe point out one example where this is observed by mentioning its corresponding repo, p_known, p_everyone and alpha values?
JamesKunstle commented on 2023-11-07T18:29:58Z ----------------------------------------------------------------
yes absolutely
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:05:59Z ----------------------------------------------------------------
Can you explain briefly what is the significance of beta and how it helps in the analysis?
JamesKunstle commented on 2023-11-07T19:09:23Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:06:00Z ----------------------------------------------------------------
maybe also print out the updated dataframe here to show the new beta column added?
JamesKunstle commented on 2023-11-07T19:10:00Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:06:01Z ----------------------------------------------------------------
Similar to Oindrilla's comment above, can you explain how to interpret the beta values and what low/high values mean?
JamesKunstle commented on 2023-11-07T19:14:26Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
hemajv commented on 2023-10-13T19:06:02Z ----------------------------------------------------------------
Similar to my comment on the previous PR, maybe include a Conclusion
section to summarize the results of the notebook?
JamesKunstle commented on 2023-11-07T19:22:49Z ----------------------------------------------------------------
++
View / edit / reply to this conversation on ReviewNB
cdolfi commented on 2023-10-26T16:55:55Z ----------------------------------------------------------------
@jameskunstle what is the pagerank score based on? Can those details be documented?
JamesKunstle commented on 2023-11-07T18:48:56Z ----------------------------------------------------------------
added in revision
I'll be rebasing this notebook only the final merged 'repo_discovery' notebook that has a longer conclusion at the end here, describing this figure and noting some ideas for next steps, including those reported below.
View entire conversation on ReviewNB
@cdolfi rebased on repo_discovery branch
rebased on top of PR #230. Extends notebook with implementations of alpha and beta terms.
New work is below section: Implement alpha and beta indices of preference
Formulation and discussion here: https://www.overleaf.com/read/hxnsqsydqcrw