arademaker

$ awk '{print $2, $2, $4, "#"}' train.pt | /media/data/src/freeling/src/utilities/train-tagger/bin/change-corpus-dict pt /usr/local/share/freeling/
O O DET # o DA0MS0 0.946534 o PD0MS000 0.0337339 o PP3MSA00 0.0197186 o NCMS000 1.38766e-05
objetivo objetivo NOUN # objetivo AQ0MS0 0.554567 objetivo NCMS000 0.390889 objetivo NP00000 0.0545441
dos dos ADP # de SPS00 1
principais principais ADJ # principal AQ0CP0 0.99061 principais NCMP000 0.00469484 principal NCMP000 0.00469484
hotéis hotéis NOUN # hotel NCMP000 1
da da ADP # de SPS00 1
cidade cidade NOUN # cidade NCFS000 1
é é VERB # ser VMIP3S0 1
que que CONJ # que PR0CN000 0.586684 que CS 0.39617 que PT0CN000 0.0167849 que PE0CN000 0.000361319

arademaker

portas <- function (t){
  teste <- 0
  amostra <- 0
  while (teste < t){
    # dist é a distribuição dos conteúdos das portas, 1 marca a porta
    # premiada
    dist <- c(1,0,0)   
    dist <- dist[sample(3)]     

    a <- sample(3,1)              # a será a porta que você escolhe    

arademaker

corpus2 <- lapply(corpus, getWords)

tira.erro <- sapply(corpus2, length) > 0
corpus2 <- corpus2[tira.erro]

apcount <- table(unlist(lapply(corpus2, names)))

apcount.df <- as.data.frame(apcount, stringsAsFactors = FALSE)
apcount.df$f <- apcount.df$Freq / length(corpus2)

arademaker

library(RCurl)
library(XML)

joinURL <- function(base, u){
  sbase <- parseURI(base)
  su <- parseURI(u)
  if( su$scheme != "")
    return(u)

  su$scheme <- sbase$scheme

arademaker

# A simple script to calculate the number of sentences per markdown
# file in a given directory. 
#
# Reference:
# - http://nltk.googlecode.com/svn/trunk/doc/howto/portuguese_en.html
# Author: Alexandre Rademaker

import os
import glob
import re

arademaker

#!/usr/bin/python
# -*- coding: utf-8 -*-

import os
import glob
import re
import codecs
import nltk
from random import *
from nltk.probability import *

arademaker

> tmp.1 <- data[grep("person",data$lex..file2,invert=TRUE),]
> subset(tmp.1, rel == "nomlex:agent", c(1,3,4,5,6,7,8,9))
           word1         word2    synset1    synset2       lex..file1     lex..file2 suffix      class
77       adaptar     adaptador 00299580-v 02678897-n      verb.change  noun.artifact     or    REGULAR
103     fabricar    fabricante 01653442-v 08060446-n    verb.creation     noun.group    nte    REGULAR
237     tripular    tripulação 01089614-v 08273167-n verb.competition     noun.group    ção    REGULAR
254      imigrar     imigração 00413432-v 08413834-n      verb.change     noun.group    ção    REGULAR
273   transferir transferência 02220461-v 01107932-n  verb.possession       noun.act    cia    REGULAR
294       servir       serviço 02541251-v 08186047-n      verb.social     noun.group      o REGRESSIVE
357         roer        roedor 01445597-v 02329401-n     verb.contact    noun.animal     or    REGULAR

arademaker

CL-USER> (pushnew "/opt/local/lib/" cffi:*foreign-library-directories* :test #'equal)
("/opt/local/lib/" (CFFI::EXPLODE-PATH-ENVIRONMENT-VARIABLE "LD_LIBRARY_PATH")
 (CFFI::EXPLODE-PATH-ENVIRONMENT-VARIABLE "DYLD_LIBRARY_PATH") (UIOP/OS:GETCWD)
 (CFFI::DARWIN-FALLBACK-LIBRARY-PATH))

CL-USER> (ql:quickload :cl-yaml)
To load "cl-yaml":
  Load 1 ASDF system:
    cl-yaml
; Loading "cl-yaml"

arademaker

;; ERC
(erc-autojoin-mode t)
(setq erc-save-buffer-on-part nil
      erc-save-queries-on-quit nil
      erc-log-write-after-send t
      erc-log-write-after-insert t
      erc-echo-notices-in-minibuffer-flag t
      erc-log-channels-directory "~/.erc/logs/"
      erc-autojoin-channels-alist '(("freenode.net" "#lisp")
				    ("wnpt.brlcloud.com" "#brl")))

arademaker

(defun group-by (alist n &optional res)
  (if (null alist)
      (reverse res)
      (group-by (subseq alist n) n (cons (subseq alist 0 n) res))))

(defun split-gloss (gloss)
  (let* ((re "[ ]*;[ ]*\"[^\"]+\"")
	 (pos (cl-ppcre:all-matches re gloss))
	 (bw '(#\; #\" #\Space)))
    (if pos