bb csv count distinct fields

count-distinct-csv.bb

(require  '[clojure.data.csv :as csv]
          '[clojure.tools.cli :refer [parse-opts]])

(def cli-options
  ;; An option with a required argument
  [["-f" "--file FILE" "File path"
    ; :default 80
    ; :parse-fn #(Integer/parseInt %)
    :validate [string? "Must be a string"]]
   ["-h" "--help"]])

(def file (:file (:options (parse-opts *command-line-args* cli-options))))


(defn csv-data->maps [csv-data]
  (map zipmap
       (->> (first csv-data) ;; First row is the header
            (map keyword) ;; Drop if you want string keys instead
            repeat)
   (rest csv-data)))



(defn read [filename]
 (csv-data->maps (csv/read-csv 
    (slurp filename) :separator \tab)))


(defn counts [filename]
 (println filename)
 (let [data (read filename)
       ks (keys (first data))]
   (pmap (juxt identity #(count (distinct (map % data)))) ks)))


(counts file)