Warning
This site has been replaced by the new QIIME 2 “amplicon distribution” documentation, as of the 2025.4 release of QIIME 2. You can still access the content from the “old docs” here for the QIIME 2 2024.10 and earlier releases, but we recommend that you transition to the new documentation at https://amplicon-docs.qiime2.org. Content on this site is no longer updated and may be out of date.
Are you looking for:
the QIIME 2 homepage? That’s https://qiime2.org.
learning resources for microbiome marker gene (i.e., amplicon) analysis? See the QIIME 2 amplicon distribution documentation.
learning resources for microbiome metagenome analysis? See the MOSHPIT documentation.
installation instructions, plugins, books, videos, workshops, or resources? See the QIIME 2 Library.
general help? See the QIIME 2 Forum.
Old content beyond this point… 👴👵
mann-whitney-u: Mann-Whitney U Test¶
| Citations |
|
|---|
Docstring:
Usage: qiime stats mann-whitney-u [OPTIONS]
Inputs:
--i-distribution ARTIFACT Dist1D[Unordered | Ordered, Independent]
[required]
--i-against-each ARTIFACT Dist1D[Unordered | Ordered, Matched |
Independent] [optional]
Parameters:
--p-compare TEXT Choices('reference', 'all-pairwise')
The comparison that will be used to analyze the
input `distribution`. Either "reference" or
"all-pairwise" must be selected. The "reference"
comparison defines Group A as the reference/control
provided to `reference-group` (sourced from either
`reference_column` or `control_column`), and Group B
as all other groups. The "all-pairwise" comparison
compares all groups to all other groups. If
`against-each` is used, this will define Group B.
[required]
--p-reference-group TEXT
If "reference" is the selected comparison, this is
the column that will be used to compare all other
groups against. [optional]
--p-alternative TEXT Choices('two-sided', 'greater', 'less')
The "two-sided" alternative hypothesis is that the
median of Group A does not equal the median of Group
B. The "greater" alternative hypothesis is that the
median of group A is greater than the median of Group
B. The "less" alternative hypothesis is that the
median of group A is less than the median of Group B.
[default: 'two-sided']
--p-p-val-approx TEXT Choices('auto', 'exact', 'asymptotic')
"exact" will calculate an exact p-value for
distributions, "asymptotic" will use a normal
distribution, and "auto" will use either "exact" when
one of the groups has less than 8 observations and
there are no ties, otherwise "asymptotic".
[default: 'auto']
Outputs:
--o-stats ARTIFACT StatsTable[Pairwise]
The Mann-Whitney U table for either the "reference"
or "all-pairwise" comparison. [required]
Miscellaneous:
--output-dir PATH Output unspecified results to a directory
--verbose / --quiet Display verbose output to stdout and/or stderr
during execution of this action. Or silence output if
execution is successful (silence is golden).
--example-data PATH Write example data and exit.
--citations Show citations and exit.
--use-cache DIRECTORY Specify the cache to be used for the intermediate
work of this action. If not provided, the default
cache under $TMP/qiime2/ will be used.
IMPORTANT FOR HPC USERS: If you are on an HPC system
and are using parallel execution it is important to
set this to a location that is globally accessible to
all nodes in the cluster.
--help Show this message and exit.
Examples:
# ### example: mann whitney pairwise
qiime stats mann-whitney-u \
--i-distribution refdist.qza \
--p-compare all-pairwise \
--i-against-each timedist.qza \
--p-p-val-approx asymptotic \
--o-stats stats.qza
Import:
from qiime2.plugins.stats.methods import mann_whitney_u
Docstring:
Mann-Whitney U Test
Parameters
----------
distribution : Dist1D[Unordered | Ordered, Independent]
compare : Str % Choices('reference', 'all-pairwise')
The comparison that will be used to analyze the input `distribution`.
Either "reference" or "all-pairwise" must be selected. The "reference"
comparison defines Group A as the reference/control provided to
`reference_group` (sourced from either `reference_column` or
`control_column`), and Group B as all other groups. The "all-pairwise"
comparison compares all groups to all other groups. If `against_each`
is used, this will define Group B.
reference_group : Str, optional
If "reference" is the selected comparison, this is the column that will
be used to compare all other groups against.
against_each : Dist1D[Unordered | Ordered, Matched | Independent], optional
alternative : Str % Choices('two-sided', 'greater', 'less'), optional
The "two-sided" alternative hypothesis is that the median of Group A
does not equal the median of Group B. The "greater" alternative
hypothesis is that the median of group A is greater than the median of
Group B. The "less" alternative hypothesis is that the median of group
A is less than the median of Group B.
p_val_approx : Str % Choices('auto', 'exact', 'asymptotic'), optional
"exact" will calculate an exact p-value for distributions, "asymptotic"
will use a normal distribution, and "auto" will use either "exact" when
one of the groups has less than 8 observations and there are no ties,
otherwise "asymptotic".
Returns
-------
stats : StatsTable[Pairwise]
The Mann-Whitney U table for either the "reference" or "all-pairwise"
comparison.