stats.t_test

Student's t-test on a numeric column — one-sample against a fixed mean, or two-sample comparing the column across the two levels of a grouping column.

Kind: standard
Network: no

Inputs🔗

Name	Type	Description
`data`	file (CSV)	Input table containing the value and (optional) grouping columns.
`value_column`	string	Name of the numeric column to test.
`group_column`	string	Optional grouping column with exactly two levels. If provided, a two-sample t-test is run; if empty, a one-sample test against `mu` is run.
`mu`	number	Null-hypothesis mean for the one-sample test. Defaults to `0`. Ignored when `group_column` is set.
`alternative`	string	Alternative hypothesis - one of `"two.sided"`, `"less"`, `"greater"`. Defaults to `"two.sided"`.
`paired`	boolean	Whether to run a paired two-sample test. Requires equal-length groups. Defaults to `false`.
`conf_level`	number	Confidence level for the reported interval. Defaults to `0.95`.

Outputs🔗

Name	Type	Description
`result`	json	Test results - statistic, degrees of freedom, p-value, confidence interval, estimate(s), and the test parameters used.

Example🔗

- id: 01890000-0000-0000-0000-000000000403
  name: stats.t_test
  inputs:
    - 01890000-0000-0000-0000-000000000000
  args:
    value_column: "biomass"
    group_column: "ownership"
    alternative: "two.sided"
    paired: false
    conf_level: 0.95