Bootstrap statistics for criterium.
Core bootstrap algorithms are provided by criterium.stats.bootstrap. This namespace provides criterium-specific integration with metrics and collect plans.
Bootstrap statistics for criterium. Core bootstrap algorithms are provided by criterium.stats.bootstrap. This namespace provides criterium-specific integration with metrics and collect plans.
Add :mean-plus-3sigma and :mean-minus-3sigma to stats map.
Add :mean-plus-3sigma and :mean-minus-3sigma to stats map.
Non-parametric BCa estimate of a statistic on data.
Non-parametric BCa estimate of a statistic on data.
Calculate bootstrap values for given estimate and samples.
Calculate bootstrap values for given estimate and samples.
Bootstrap a statistic with BCa confidence intervals.
Bootstrap a statistic with BCa confidence intervals.
Mean, variance and confidence interval.
Mean, variance and confidence interval.
Bootstrap sampling of a statistic, using resampling with replacement.
Bootstrap sampling of a statistic, using resampling with replacement.
(bootstrap-stats)(bootstrap-stats {:keys [id metric-ids samples-id outliers-id ess-id]
:as analysis})Analysis function that adds bootstrap stats to the result.
Parameters: opts - Optional map with keys: :id - Key for bootstrap stats in output (default: :bootstrap-stats) :samples-id - Key for source samples (default: :samples) :outliers-id - Key for outlier analysis if available (default: :outliers)
:ess-id - Key for effective sample size analysis (default: nil).
When provided and ESS data exists, CI widths are inflated
by the ci-inflation-factor and stored as
:adjusted-estimate-quantiles.
:metric-ids - Set of metric ids to analyze (default: all quantitative)
:quantiles - Additional quantiles to compute (default: none)
:estimate-quantiles - Confidence interval quantiles
(default: [0.025 0.975])
:bootstrap-size - Number of bootstrap resamples
(default: sample count)
:min-samples - Minimum sample size for reliable results
(default: 30).
When sample count is below this threshold, a warning
is printed and results include
:low-sample-count? true.
When :outliers-id is provided, outliers identified in the outlier analysis are removed from samples before bootstrap resampling. This prevents outliers from propagating and amplifying in resamples.
When :ess-id is provided and the data-map contains effective sample size analysis under that key, confidence intervals are inflated by the ci-inflation-factor for each metric. The adjusted CIs are stored as :adjusted-estimate-quantiles alongside the original :estimate-quantiles. If ESS data is not present for a metric, only the original CIs are included.
Analysis function that adds bootstrap stats to the result.
Parameters:
opts - Optional map with keys:
:id - Key for bootstrap stats in output
(default: :bootstrap-stats)
:samples-id - Key for source samples (default: :samples)
:outliers-id - Key for outlier analysis if available
(default: :outliers)
:ess-id - Key for effective sample size analysis (default: nil).
When provided and ESS data exists, CI widths are inflated
by the ci-inflation-factor and stored as
:adjusted-estimate-quantiles.
:metric-ids - Set of metric ids to analyze (default: all quantitative)
:quantiles - Additional quantiles to compute (default: none)
:estimate-quantiles - Confidence interval quantiles
(default: [0.025 0.975])
:bootstrap-size - Number of bootstrap resamples
(default: sample count)
:min-samples - Minimum sample size for reliable results
(default: 30).
When sample count is below this threshold, a warning
is printed and results include
:low-sample-count? true.
When :outliers-id is provided, outliers identified in the outlier analysis
are removed from samples before bootstrap resampling. This prevents outliers
from propagating and amplifying in resamples.
When :ess-id is provided and the data-map contains effective sample size
analysis under that key, confidence intervals are inflated by the
ci-inflation-factor for each metric. The adjusted CIs are stored as
:adjusted-estimate-quantiles alongside the original :estimate-quantiles.
If ESS data is not present for a metric, only the original CIs are included.(bootstrap-stats* metric->values outliers metric-configs config)Compute bootstrap stats for all metric paths. When outliers is non-nil, removes outlier samples before bootstrap resampling.
Compute bootstrap stats for all metric paths. When outliers is non-nil, removes outlier samples before bootstrap resampling.
(bootstrap-stats-for samples opts)Compute bootstrap statistics for samples with given options.
Accepts DoubleArray or vector of samples. Computes mean, variance, and quantiles with BCa confidence intervals. Does not include min-val, max-val, or 3-sigma bounds.
Raw values are returned without transform application. Transforms are applied later via the source-id chain when viewing results.
Options: :bootstrap-size - Number of bootstrap resamples (default: sample count) :min-samples - Minimum sample size threshold (default: 30). When sample count is below this, a warning is printed and results include :low-sample-count? true.
The :bootstrap-size option controls the number of bootstrap resamples. Defaults to the number of samples if not specified.
Compute bootstrap statistics for samples with given options.
Accepts DoubleArray or vector of samples.
Computes mean, variance, and quantiles with BCa confidence intervals.
Does not include min-val, max-val, or 3-sigma bounds.
Raw values are returned without transform application. Transforms are
applied later via the source-id chain when viewing results.
Options:
:bootstrap-size - Number of bootstrap resamples (default: sample count)
:min-samples - Minimum sample size threshold (default: 30). When sample
count is below this, a warning is printed and results
include :low-sample-count? true.
The :bootstrap-size option controls the number of bootstrap resamples.
Defaults to the number of samples if not specified.Return coll with element at index n removed.
Return coll with element at index n removed.
Scale a BcaEstimate by the given scale factor.
Scale a BcaEstimate by the given scale factor.
Apply function f to all values in stats map.
Apply function f to all values in stats map.
Combine multiple stat functions into one.
Combine multiple stat functions into one.
Map of stat keywords to their corresponding functions.
Map of stat keywords to their corresponding functions.
Build vector of stat functions including quantile functions.
Build vector of stat functions including quantile functions.
cljdoc builds & hosts documentation for Clojure/Script libraries
| Ctrl+k | Jump to recent docs |
| ← | Move to previous article |
| → | Move to next article |
| Ctrl+/ | Jump to the search field |