Created
December 3, 2009 16:43
-
-
Save angerman/248320 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
;; LibSVM Model Parser | |
(use 'clojure.contrib.duck-streams) | |
(use 'clojure.contrib.seq-utils) | |
(require '[clojure.contrib.str-utils2 :as s]) | |
(load "coll-util") | |
(def *root* "/Users/angerman/Downloads/libsvm-2.9-dense/tools/") | |
;; (set! *warn-on-reflection* true) eval in repl | |
(defn model-file [name] (str *root* name)) | |
(defn str-to-int [s] (Integer/parseInt s)) | |
(defn str-to-float [s] (Float/parseFloat s)) | |
(defn str-to-double [s] (Double/parseDouble s)) | |
;; turn "key_word value" into {:key-word value} | |
(defn parse-header-line [line] | |
(let [[_ key val] (re-matches #"^([^ ]*) (.*)$" line) | |
key (keyword (s/replace key "_" "-"))] | |
{key val})) | |
;; parse an <feature:int>:<value:float> item into {feature value} | |
(defn parse-sparse-item [item] | |
(let [[feature value] (s/split item #":")] | |
{(str-to-int feature) (str-to-float value)})) | |
;; SV lines look like [class1-vs-class2 ... class1-vs.classN <feature>:<value>*] | |
;; where feature is an int and value a float | |
(defn process-sv-line [classes line] | |
(let [nr-alpha (dec classes) | |
items (s/split line #" ") | |
;; alpha values are stored as doubles | |
alpha (map str-to-double (take nr-alpha items)) | |
sv (into {} (map parse-sparse-item (drop nr-alpha items)))] | |
(list alpha sv))) | |
(defn rewrite-header [header] | |
"prettifyes the header map" | |
(let [nr-sv (map str-to-int (s/split (:nr-sv header) #" "))] | |
(assoc header | |
:nr-class (str-to-int (:nr-class header)) | |
:total-sv (str-to-int (:total-sv header)) | |
:rho (map str-to-float (s/split (:rho header) #" ")) | |
:label (map str-to-int (s/split (:label header) #" ")) | |
:nr-sv nr-sv | |
:offsets (cons 0 (reductions + (butlast nr-sv)))))) | |
(defn parse-header [file] | |
(loop [header {} | |
lines (read-lines file)] | |
(let [line (first lines)] | |
(if (re-matches #"^SV$" line) | |
[(rewrite-header header) | |
(next lines)] | |
(recur (into header (parse-header-line line)) | |
(next lines)))))) | |
(defn drop-header [file] | |
(loop [lines (read-lines file)] | |
(let [line (first lines)] | |
(if (re-matches #"^SV$" line) | |
(next lines) | |
(recur (next lines)))))) | |
(defn extract-features [line] | |
(map (comp str-to-int second) (re-seq #"(\d+):" line))) | |
(defn compute-metrics [[header lines]] | |
(assoc header :features (reduce combine-sorted (map extract-features lines)))) | |
; (loop [[ln & lns*] lines | |
; features nil | |
; experiments 0] | |
; (printf "%d, %d\n" experiments (count features)) | |
; (if (not ln) | |
; (assoc header | |
; :experiments experiments | |
; :features features) | |
; (recur lns* | |
; (time (combine-sorted features (extract-features ln))) | |
; (inc experiments))))) | |
;; reads the model and binds it to *model* | |
(def *model* (compute-metrics (parse-header (model-file "upps.data.model")))) | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment