Metrics reference

Each metric is characterised by the following traits:

  • A name, together with a path in metrics tree (e. g. "Halstead.Volume" means the name of a metric is "Volume" and it is in the "Halstead" section).
  • One of two types: base or composite.
    • The value of a base metric is calculated based on the source code of a function (e. g. Lines of Code).
    • The value of a composite metric is calculated based only on the values of other metrics. (e. g. Halstead.Length = Hasltead.N1 + Halstead.N2) Metric's dependencies can be both base and composite.
  • "Good" value, being the target for this metric. If a metric of a function has this value, it is colored in green.
  • "Too low" and/or "Too high" values. If a metric of a function has any of these values, it is colored in red. Colors of values between "Good" and "Too low"/"Too high" are interpolated. Values further from "Good" than "Too low"/"Too bad" are colored red.

The "Good", "Too low" and "Too bad" values can be set in the interface. See Histograms.

Cyclomatic complexity

Type: base

Available in: JavaCPython

Number of independent execution paths through the function. Calculated as 1 + number of decision points. Decision points are branches, loops, and logical operators that can alter the flow of execution.

Halstead metrics

n1

Type: base

Available in: JavaCPython

Number of distinct operators used in the function. Operators are language-specific and can include arithmetic operators, logical operators, assignment operators, etc.

n2

Type: base

Available in: JavaCPython

Number of distinct operands used in the function. Operands are language-specific and can include variable names, literals, function calls, etc.

N1

Type: base

Available in: JavaCPython

Number of total operators used in the function. Operators are language-specific and can include arithmetic operators, logical operators, assignment operators, etc.

N2

Type: base

Available in: JavaCPython

Number of total operands used in the function. Operands are language-specific and can include variable names, literals, function calls, etc.

Vocabulary

Type: composite

The total number of unique operators and operands in the function (n1 + n2).

Length

Type: composite

The total number of operators and operands in the function (N1 + N2).

Estimated length

Type: composite

Estimate of the length of the program based on its vocabulary: n1 * log2(n1) + n2 * log2(n2)

Volume

Type: composite

Total information of the program. N * log2(n), where N is the program length and n is the program vocabulary.

Difficulty

Type: composite

Difficulty of the program to write or understand. (n1 / 2) * (N2 / n2).

Effort

Type: composite

The effort required to write or understand the program. The product of the program difficulty and volume.

Time

Type: composite

The estimated time required to write the program. The effort divided by 18 seconds.

Bugs

Type: composite

The estimated number of bugs delivered in the code. The effort raised to the power of 2/3 divided by 3000.

Lines of Code

Available in: JavaCPython

Type: base

Number of lines in the function, including signature, blank lines, comments and the trailing brace.