understand.txt

Results obtained from understand are collected in table
understand_raw -- they are not available as ready-made plots,
but are neither in complete raw form since the amount of information
obained from understand is impractical to analyse quickly enough
for the web frontend.

After applying understand to sample checkouts of a project
(typically between 5-15 per release range), the quantile 
values of the distribution are computed for each measurement
(for instance, Cyclomatic Complexity, Knots,  ...) are
computed for each artefact (class, file, function, ...) that can
be analysed with understand (not all combinations are valid).

The structure of table understand_raw is as follows:

+----------+-------------+------+-----+---------+-------+
| Field    | Type        | Null | Key | Default | Extra |
+----------+-------------+------+-----+---------+-------+
| plotId   | bigint(20)  | NO   | MUL | NULL    |       |
| time     | datetime    | NO   |     | NULL    |       |
| kind     | varchar(30) | NO   | MUL | NULL    |       |
| name     | varchar(45) | YES  |     | NULL    |       |
| variable | varchar(45) | NO   |     | NULL    |       |
| value    | double      | NO   |     | NULL    |       |
+----------+-------------+------+-----+---------+-------+


* kind is the type of artefact analysed (class, file, function, ...)
* name can assume the following values:
  - q1, q2, q3, q4, q5: extreme of the lower whisker, the lower ‘hinge’,
                        the median, the upper ‘hinge’ and the extreme of
                        the upper whisker (see boxplot.stat)
  - num.observations: Number of, well, observations recorded by understand
    and used to compute the summary statistics
  - outlier: Outlier value (outliers are recorded one-by-one; the other
    values are obviously available only once)
  - conf1 and conf2: lower and upper extremes of the ‘notch’ (see
    boxplot.stats)
* variable denotes the type of measurement: RatioCommentToCode, MaxNesting,
  CountPath, ...

Time series can be generated from the raw data: To see how the
median of the ratio of code and comments evolved over time, the following
query needs to be run:

SELECT time, value FROM understand_raw \
       WHERE variable="RatioCommentToCode" AND kind="File" AND name="q3";