fuzzy comparing floating point numbers

fuzzy_compare_floating_point_numbers.clj

;; This turns out to be very, very useful when working with floating point numbers in some cases.
;; E.g. if you switch some of your input from ^long to ^double your calculations might actually shift ever so slightly
;; which will make comparing old vs. new results tricky.

;; You'll need this in your project.clj file:

[com.google.guava/guava "LATEST"]


;; ..and this somewhere useful:

(import 'com.google.common.math.DoubleMath)


;; Now instead of comparing using =, you can do something like:

quantataraxia.core> (let [md (oneb-get-ohlc-with-spreads "EURUSD" (tp-seconds (time/hours 1)))]
                      (dorun (map #(try (assert (DoubleMath/fuzzyEquals %1 %2 0.000000000000001))
                                        (catch Throwable e
                                          (println [%1 %2])
                                          #_(throw e)))
                                  (old-wma (double-array (map pclose md)) 15)
                                  (wma (double-array (map pclose md)) 15))))