Expand description
Shared helpers for the statistics module.
All numeric conversions are explicit: float64 never silently enters exact arithmetic, non-finite values are rejected, and non-numeric array entries are reported with their path.
Enums§
- Moment
- An exact or approximate statistical moment.
- Series
- A numeric series in the representation chosen for the calculation.
Functions§
- all_
modes - Descriptive statistics support all three numeric modes.
- allocation_
ratio_ param - alpha_
param - any_
number_ schema - array_
schema - array_
value - assumptions_
value - A text array of method assumptions, shared by the inference records.
- bool_
schema - bool_
value - classify_
series - Convert an input array to either exact rationals or float64 values.
- collect_
numbers - compensated_
sum - Neumaier compensated summation.
- confidence_
param - covariance_
exact - covariance_
f64 - Two-pass centered covariance with compensated summation.
- covariance_
moment - ensure_
finite - example_
args - extreme_
number - field
- float64_
schema - float_
array - float_
number - float_
value - float_
vector - Collect a non-empty numeric array as float64 values, preserving the path of the offending entry in any conversion error.
- inferential_
modes - Distributions and inference always produce float64 results and therefore require auto or scientific mode.
- insufficient
- integer_
schema - integer_
value - mean_
exact - mean_
f64 - mean_
moment - median_
number - mode_
numbers - All tied modes and their counts, sorted ascending.
- non_
negative_ u64 - number_
field - number_
to_ f64 - number_
value - optional_
f64_ param - optional_
field - parse_
choice - parse_
value - power_
param - probability_
param - quantile_
number - R-7 style quantiles with alternative order-statistic conventions.
- rational_
sqrt_ exact - Exact square root of a non-negative rational, when one exists.
- rational_
to_ number - rational_
value - record
- record_
schema - scalar_
f64 - sided_
param - stddev_
moment - Convert a variance moment to a standard deviation under the mode rules: an exact perfect square stays exact, otherwise exact mode is rejected, auto mode returns a decimal approximation, and scientific mode returns float64.
- text
- text_
schema - variance_
exact - variance_
f64 - Welford’s online algorithm for the sample variance.
- variance_
moment