Slow down decay rate of enclosing methods

ponder-lab / Rejuvenate-Logging-Levels

An Eclipse plug-in research prototype for automatically rejuvenating log statement levels using degree of interest

Eclipse Public License 2.0

2 stars 3 forks source link

Slow down decay rate of enclosing methods #232

Closed khatchad closed 5 years ago

khatchad commented 5 years ago

Highly related to #224. In fact, this should have been fixed since #224 was merged. What happened? If #224 was merged, why do we have this problem still (or do we not)? What was the new decay rate there?

The decay rate may have to be tied to the project being analyzed, e.g., number of commits, SLOC.

khatchad commented 5 years ago

Highly related to #217.

yiming-tang-cs commented 5 years ago

original decay /64 results_decay_64.zip

original decay: results_original_decay.zip

commits:24000 I tried 1000, but the number of candidates is small. 24000 is much better.

I will try more decays.

yiming-tang-cs commented 5 years ago

original decay/16: results_div_16.zip

original decay/32: results_div_32.zip

original decay/128: results_div_128.zip

original decay/256: results_div_256.zip

khatchad commented 5 years ago

Please try 16 and 256 on:

[ ] One medium subject.
[ ] One small subject.

yiming-tang-cs commented 5 years ago

I tried one small subject: reactiondecoder, but there is no candidate in the results. I need to find another one.

Reaction_Decoder.zip

yiming-tang-cs commented 5 years ago

This result below is interesting. So many candidates but no transformation. results_IRCT_API.zip

yiming-tang-cs commented 5 years ago

By looking at results for IRCT_API, I think we need to fix #236 before doing this. IRCT-API is a good small project for this issue.

khatchad commented 5 years ago

I don't think it makes any sense to work on this issue with #236 pending.

khatchad commented 5 years ago

Blocked on #236.