Skip to content
Docs Pydantic Docs

statistics module

mean, fmean, median, median_low, median_high, mode, multimode, pstdev, pvariance, stdev, variance, quantiles and StatisticsError are available.

  • NormalDist, kde, kde_random, geometric_mean, harmonic_mean, median_grouped, covariance, correlation and linear_regression are not provided. They are missing from the module namespace, so they fail type checking as well as raising AttributeError at runtime.
  • Decimal and Fraction data are not supported by mean, variance, pvariance, stdev and pstdev. Monty has no fractions module, and these functions raise TypeError: can't convert type '<type>' to numerator/denominator, the error CPython gives for a type with no as_integer_ratio(). median_low, median_high and mode work on any orderable (or hashable) value, as in CPython; median and quantiles also need data that can be added and divided.
  • mode() and multimode() of unhashable data raise Counter’s message, which differs from CPython’s (see collections.md).
  • StatisticsError reprs as statistics.StatisticsError(...), as the other dotted exception classes do (see exceptions.md).
  • quantiles() with an n too large to allocate raises MemoryError before computing any cut point, where CPython appends cut points until it runs out of memory.
  • The module has no __all__, and its functions carry none of the attributes CPython’s Python functions have, as on every built-in module (see modules.md).