Update references

README.md

@@ -75,7 +75,12 @@ Which in this context means: let's not focus on benchmarks and performance, but
 ## References
 * [numpy.random](https://numpy.org/devdocs/reference/random/generator.html): distributions and random sampling
 * [numpy statistics](https://numpy.org/devdocs/reference/routines.statistics.html#averages-and-variances): order statistics, averages and variances
+* [scipy stats](https://github.com/scipy/scipy/blob/3de0d58/scipy/stats/stats.py) module and related [tests](https://github.com/scipy/scipy/blob/1150c4c033899a5a4556b7d34d6b137352b36b9e/scipy/stats/tests/test_stats.py) tests
+* [julia random](https://docs.julialang.org/en/v1/stdlib/Random/) module
 * [julia statistics](https://docs.julialang.org/en/v1/stdlib/Statistics/#Statistics.std) module
+* [julia distributions](https://juliastats.org/Distributions.jl/latest/starting/) package.
+* on [skewness and kurtosis](https://brownmath.com/stat/shape.htm), by Stan Brown
+* more on [skewness and kurtosis](https://www.itl.nist.gov/div898/handbook/eda/section3/eda35b.htm), from NIST.
 
 ## Contributing
 

src/statistics.cr

@@ -119,6 +119,8 @@ module Statistics
   # - `values`: a one-dimensional dataset.
   # - `p`: probability. Values of `p` should be in the interval `[0, 1]`.
   # - `sorted` indicates whether `values` can be assumed to be sorted.
+  #
+  # Implementation based on Julia's [Statistics.quantile](https://docs.julialang.org/en/v1/stdlib/Statistics/#Statistics.quantile).
   def quantile(values, p, sorted = false)
     sorted_values = sorted ? values : values.sort
     n = values.size