Hi, thank you for your incredibly good work. I have a question.
MAR and MFR of DepGraph are less than 1 for project JacksonXml in table2. Can you explain?
In your paper MAR is defined as usual which is "Mean Average Rank (MAR). For each faulty version, the Mean Average Rank (MAR) measures the average position of all faulty methods in a list. For each project, MAR fnds the average of these
positions across all its faulty versions. The lower the value, the better the ranking."
MFR is "Mean First Rank (MFR). This metric determines the position of the frst identifed fault in the ranked list for every faulty version. The MFR for a project is then derived by taking the average of these positions across all of its faulty versions. A lower MFR means the faulty methods can be identifed earlier in the list."
In my opinion, both of MAR and MFR must be bigger than 1 in FL task.
Can you explain how these numbers were calculated?
Thank you.
wuhecuizao
commented
4 days ago 
Hi, thank you for your incredibly good work. I have a question. MAR and MFR of DepGraph are less than 1 for project JacksonXml in table2. Can you explain?
In your paper MAR is defined as usual which is "Mean Average Rank (MAR). For each faulty version, the Mean Average Rank (MAR) measures the average position of all faulty methods in a list. For each project, MAR fnds the average of these positions across all its faulty versions. The lower the value, the better the ranking." MFR is "Mean First Rank (MFR). This metric determines the position of the frst identifed fault in the ranked list for every faulty version. The MFR for a project is then derived by taking the average of these positions across all of its faulty versions. A lower MFR means the faulty methods can be identifed earlier in the list."
In my opinion, both of MAR and MFR must be bigger than 1 in FL task. Can you explain how these numbers were calculated? Thank you.