Estimation of Changes to Annotated Functions Depending on Snapshot Size

Data to estimate the amount of changes to annotated functions from one snapshot to the next, depending on snapshot size.

Snapshot sizes: 50, 100, 200, 400
System: openldap

Commands Used

For instance:

[me@mymachine: ~/src/skunk/snapshotsize-050]
$ stat-function-changes-between-windows.sh --all openldap > \
>  results/openldap/changed*functions.csv
cd ../snapshotsize-100
[me@mymachine: ~/src/skunk/snapshotsize-100]
$ stat-function-changes-between-windows.sh --all openldap > \
>  results/openldap/changed*functions.csv
...
[me@mymachine: ~/src/skunk/snapshotsize-400]
$ stat-function-changes-between-windows.sh --all openldap > \
>  results/openldap/changed*functions.csv
cd ..
$ for d in snapshotsize-*; do
>   echo "*** $d ***";
>   csvstat $d/results/openldap/changed_functions.csv;
> done|g -F -e '***' -e FUNC -e Mean -e Median  |\
>   sed 's/^[[:space:]]*[0-9]\{1,2\}\./*/'

Results for Annotated Functions

Column Explanation

AB_BEFORE: annotated functions existing in previous snapshot
AB_NOW: annotated functions existing in current snapshot
AB_IDENTICAL: annotated functions existing in both snapshots
AB_ADDED: annotated functions added compared to previous snapshot
AB_REMOVED: annotated functions removed compared to previous snapshot

Metrics about number of annotated functions existing in both snapshots whose annotation metric changed:

AB_CH_NOFL: NOFL changed
AB_CH_NOFC_NONDUP: NOFC_NONDUP changed
AB_CH_NONEST: NONEST changed

Changes expressed as percentages:

AB_PERC_ADD_REM: Percentage of added or removed annotated functions, compared to AB_BEFORE

Percentage of annotated functions that …

underwent at least one commit,
exist in both snapshots
whose annotation metric changed.

AB_PERC_CH_NOFL: % of annotated, changed functions where NOFL changed
AB_PERC_CH_NOFC_NONDUP: % of annotated, changed functions where NOFC_NONDUP changed
AB_PERC_CH_NONEST: % of annotated, changed functions where NONEST changed

50 Commits

Data in CSV format

AB_BEFORE
- Mean: 619.32
- Median: 586.0
AB_NOW
- Mean: 620.90
- Median: 586.5
AB_IDENTICAL
- Mean: 608.88
- Median: 582.0
AB_ADDED
- Mean: 12.02
- Median: 7.0
AB_REMOVED
- Mean: 10.43
- Median: 4.0
AB_CH_NOFL
- Mean: 7.42
- Median: 5.0
AB_CH_NOFC_NONDUP
- Mean: 5.49
- Median: 3.0
AB_CH_NONEST
- Mean: 1.93
- Median: 1.0
AB_PERC_ADD_REM
- Mean: 3.73
- Median: 2.0
AB_PERC_CH_NOFL
- Mean: 13.19
- Median: 11.0
AB_PERC_CH_NOFC_NONDUP
- Mean: 9.19
- Median: 6.0
AB_PERC_CH_NONEST
- Mean: 2.99
- Median: 1.0

100 Commits

Data in CSV format

AB_BEFORE
- Mean: 618.65
- Median: 585.5
AB_NOW
- Mean: 621.76
- Median: 586.5
AB_IDENTICAL
- Mean: 598.56
- Median: 574.5
AB_ADDED
- Mean: 23.20
- Median: 16.0
AB_REMOVED
- Mean: 20.09
- Median: 9.5
AB_CH_NOFL
- Mean: 13.35
- Median: 10.0
AB_CH_NOFC_NONDUP
- Mean: 10.08
- Median: 7.0
AB_CH_NONEST
- Mean: 3.60
- Median: 2.0
AB_PERC_ADD_REM
- Mean: 7.70
- Median: 4.5
AB_PERC_CH_NOFL
- Mean: 14.32
- Median: 13.0
AB_PERC_CH_NOFC_NONDUP
- Mean: 10.32
- Median: 9.0
AB_PERC_CH_NONEST
- Mean: 3.38
- Median: 2.0

200 Commits

Data in CSV format

AB_BEFORE
- Mean: 617.80
- Median: 586
AB_NOW
- Mean: 624.10
- Median: 588
AB_IDENTICAL
- Mean: 579.61
- Median: 566
AB_ADDED
- Mean: 44.49
- Median: 37
AB_REMOVED
- Mean: 38.20
- Median: 22
AB_CH_NOFL
- Mean: 22.86
- Median: 17
AB_CH_NOFC_NONDUP
- Mean: 17.62
- Median: 13
AB_CH_NONEST
- Mean: 6.52
- Median: 5
AB_PERC_ADD_REM
- Mean: 15.14
- Median: 10
AB_PERC_CH_NOFL
- Mean: 15.68
- Median: 15
AB_PERC_CH_NOFC_NONDUP
- Mean: 11.75
- Median: 10
AB_PERC_CH_NONEST
- Mean: 3.99
- Median: 3

400 Commits

Data in CSV format

AB_BEFORE
- Mean: 613.66
- Median: 583
AB_NOW
- Mean: 625.89
- Median: 588
AB_IDENTICAL
- Mean: 543.57
- Median: 548
AB_ADDED
- Mean: 82.31
- Median: 64
AB_REMOVED
- Mean: 70.09
- Median: 52
AB_CH_NOFL
- Mean: 37.29
- Median: 27
AB_CH_NOFC_NONDUP
- Mean: 29.26
- Median: 22
AB_CH_NONEST
- Mean: 10.54
- Median: 7
AB_PERC_ADD_REM
- Mean: 28.71
- Median: 20
AB_PERC_CH_NOFL
- Mean: 17.37
- Median: 17
AB_PERC_CH_NOFC_NONDUP
- Mean: 13.20
- Median: 12
AB_PERC_CH_NONEST
- Mean: 4.40
- Median: 4

Results for All Functions

FUNC_BEFORE: functions existing in previous snapshot
FUNC_NOW: functions existing in current snapshot
FUNC_IDENTICAL: functions existing in both snapshots
FUNC_ADDED: functions added compared to previous snapshot
FUNC_REMOVED: functions removed compared to previous snapshot

Changes expressed as percentages:

FUNC_PERC_ADD_REM: Percentage of added or removed functions, compared to FUNC_BEFORE
FUNC_MEDIAN_PERC_LOC_CHG: Median value of the percentage by which the LOC metric of a function has changed, given that the function received at least one commit.