pastk/organicmaps
1
Fork 0
forked from organicmaps/organicmaps
Code Pull requests Activity

Merge pull request #2937 from ygorshenin/fixed-postcodes-clusterization-script

Browse source 
[search] Fixes to postcodes clusterization script.
This commit is contained in:
mgsergio 2016-04-21 15:06:58 +04:00
commit bec3cbc551
1 changed files with 15 additions and 10 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

25
search/search_quality/clusterize-postcodes.lisp
View file

@ -28,15 +28,19 @@ exec /usr/bin/env sbcl --noinform --quit --load "$0" --end-toplevel-options "$@"
(string-trim *seps* string))
(defun get-postcode-pattern (postcode)
"Simplifies postcode in a following way:
* all letters are replaced by 'a'
* all digits are replaced by '0'
* other characters are left as-is
"Simplifies postcode in the following way:
* all letters are replaced by 'A'
* all digits are replaced by 'N'
* hyphens and dots are replaced by a space
* other characters are capitalized
This format follows https://en.wikipedia.org/wiki/List_of_postal_codes.
"
(map 'string #'(lambda (c) (cond ((alpha-char-p c) #\a)
((digit-char-p c) #\0)
(map 'string #'(lambda (c) (cond ((alpha-char-p c) #\A)
((digit-char-p c) #\N)
((or (char= #\- c) (char= #\. c)) #\Space)
(T c)))
postcode))
(string-upcase postcode)))
(defun get-pattern-clusters (postcodes)
"Constructs a list of clusters by a list of postcodes."
@ -89,8 +93,9 @@ exec /usr/bin/env sbcl --noinform --quit --load "$0" --end-toplevel-options "$@"
; Prints number of postcodes in a cluster, accumulated
; percent of postcodes clustered so far, simplified version
; of a postcode and examples of postcodes.
(format t "~a (~2$%) ~a [~{~a~^, ~}]~%"
(format t "~a (~2$%) ~a [~{~a~^, ~}~:[~;, ...~]]~%"
num-samples
(coerce (* 100 (/ curr-prefix-sum *total*)) 'double-float)
(* 100 (/ curr-prefix-sum *total*))
key
(subseq samples 0 (min num-samples 5)))))
(subseq samples 0 (min num-samples 5))
(> num-samples 5))))