correlation-matrix.clj

(ns behrica.corr-matrix
  (:require [fastmath.stats :as stats]
            [aerial.hanami.common :as hc]
            [aerial.hanami.templates :as ht]
            [aerial.hanami.core :as hmi]))

(defn round
  [n scale rm]
  (.setScale ^java.math.BigDecimal (bigdec n)
             (int scale)
             ^RoundingMode (if (instance? java.math.RoundingMode rm)
                             rm
                             (java.math.RoundingMode/valueOf
                              (str (if (ident? rm) (symbol rm) rm))))))

(defn calc-correlations-matrix [data cols-to-use]
  (doall
   (for [col-1 cols-to-use col-2 cols-to-use]
     {:col-1 col-1
      :col-2 col-2
      :corr
      (Float/valueOf
       (str
        (round
         (stats/pearson-correlation (data col-1) (data col-2))
         2 :DOWN)))})))





(def text-encoding
  (-> ht/xy-encoding
      (assoc :text
             {:field :CORR
              :type "quantitative"}
             :color {:condition
                     {"test" {:field :CORR :lt 0} "value" "black"}
                     "value" "white"})
      (dissoc :x :y :tooltip)))

(def text-layer
  (-> ht/view-base
    (assoc :mark (merge ht/mark-base {:type "text"})
           :encoding text-encoding)
    (dissoc :usermeta :data :height :width)))

(def rect-layer
  (-> ht/view-base
    (assoc :mark (merge ht/mark-base {:type "rect"})
           :encoding (-> ht/xy-encoding
                         (dissoc :x :y :tooltip)
                         (assoc :color {:field :CORR :type "quantitative"
                                        :scale {:scheme "redgrey" :reverse true}})))
    (dissoc :usermeta :data :height :width)))

(def heatmap-chart2
  (assoc ht/layer-chart
         :encoding ht/xy-encoding
         :layer [rect-layer text-layer]))


(def heatmap-chart
  {
   :data :UDATA
   :encoding :ENCODING
   :height :HEIGHT :width :WIDTH
   :layer [{
            :mark "rect"
            :encoding
            {
             :color {:field :CORR-COL :type :quantitative
                     :scale {:scheme "redgrey" :reverse true}}}}

           {:mark "text",
            :encoding
            {:text {:field :CORR-COL :type "quantitative"}
             :color
             {:condition {"test" {:field :CORR-COL :lt 0} "value" "green"},
              "value" "blue"}}}]})

(def correlation-chart
  {
   :data :UDATA
   :height :HEIGHT :width :WIDTH
   :layer [{
            :mark "point"
            :encoding
            {:x { :field :COL-NAME :type :quantitative}
             :y { :field :BASE-COL-NAME :type :quantitative}}}

           {:mark {:type :line :color :firebrick}
            :transform
            [{:regression :BASE-COL-NAME :on :COL-NAME}]
            :encoding
            {:x {:field :COL-NAME :type :quantitative}
             :y {:field :BASE-COL-NAME :type :quantitative}}}

           {:transform
            [{:regression :BASE-COL-NAME
              :on :COL-NAME
              :params true}
             {:calculate "'R²: '+format(datum.rSquared, '.2f')", :as "R2"}],
            :mark
            {:type "text", :color "firebrick", :x "width", :align "right", :y -5},
            :encoding {:text {:type "nominal", :field "R2"}}}]})