Problem with #185

[khatchad]

There is a problem with the change made per #185. Basically, we are missing the coverage of the DOI spectrum because we removed the last two intervals. What happens when a logging statement's enclosing method DOI value falls outside of the intervals in the boundaries?

[khatchad]

Things to try:

• [x] Try reverting the change made for #185 because now we have a different decay rate. This may help alleviate the problem we were having before.
• [x] Try adding the rest of the spectrum to the last bucket. That last one (INFO) becomes very large.

[khatchad]

For the first one, just remove the IntStream loop.

[yiming-tang-cs]

This the result after reverting the change:

results_revert.zip

[yiming-tang-cs]

This is the results after modifying the last bucket:

results_last_boundary.zip

[khatchad]

• [x] Try another benchmark.
• [x] Try another setting.

[yiming-tang-cs]

I've tried another benchmark IRCT-API.

[yiming-tang-cs]

IRCT-API with option 1 (reverting change): irct_api_log_option_1.tar.gz

IRCT-API with option 2 (adjusting last bucket): irct_api_log_option2.tar.gz

[yiming-tang-cs]

Selenium with option 1 + 6 settings: selenium_log_option_1.zip

[yiming-tang-cs]

IRCT-API option 1 without log scale: IRCT_API_option_1_without_log_scale.zip

IRCT-API option 2 without log scale: IRCT_API_log_option_2_without_log_scale.zip

[khatchad]

We will go with option 1, effectively reverting #185.