Median absolute deviation
Median absolute deviation is a robust measure of spread equal to the median of observations' absolute deviations from their median.
Outlier Detection in Python
Median absolute deviation is a robust measure of statistical spread equal to the median of the absolute deviations from the data's median.
- How it is used
- Detects univariate numeric outliers by comparing each observation’s deviation from the median with the typical deviation.
- How deep it goes
- Explains calculation, scoring, thresholds, Python implementation, and aggregation limits without developing a broader multivariate detector.
- Examples
- Uses abalone whole-weight data, an inserted value of 4.0, categorical counts, and segment features to illustrate scoring behavior.
- What you can do after
- Apply it via Python to calculate robust deviation scores and flag candidate outliers using a chosen threshold.
Read how Outlier Detection in Python explains median absolute deviation →
Statistics Slam Dunk
Median absolute deviation is a measure of dispersion based on the absolute differences between observations and the median of the data set.
- How it is used
- Measures within-season NBA win dispersion to evaluate whether salary-cap-era parity improved.
- How deep it goes
- Defines the measure, implements grouped R arithmetic, and interprets a trend chart alongside other dispersion measures.
- Examples
- Uses NBA team wins grouped by season, R tibble columns, an annual line chart, and pre- versus post-salary-cap comparisons.
- What you can do after
- Apply it via R to calculate team-level median deviations and compare dispersion across seasons.
Read how Statistics Slam Dunk explains median absolute deviation →
Related concepts
- mean absolute deviationalternative to · Mean absolute deviation is the average absolute distance of observations from a specified reference value, commonly their mean.
- countuses · A count is a numerical tally or operation determining how many items, occurrences, records, or values satisfy specified criteria.
- variancekind of · Variance is the expected squared deviation of a random variable from its mean, measuring the spread of its distribution.
- thresholdconfigures · A threshold is a boundary value against which a quantity is compared to determine whether a condition is met or an action or decision is triggered.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.