jelsas

#!/bin/sh

#
# A Twitter stream sucker. Writes one file per hour, containing all the tweets from the
# sample stream (aka "gardenhose").
#
# Use Ctrl-C to stop collecting.
#
# Note: This is not intended to be a robust tool for collecting Twitter data. Use at your 
# own risk, and play nice: http://dev.twitter.com/pages/streaming_api_concepts

jelsas

# the following two lines give a two-line status, with the current window highlighted
#hardstatus alwayslastline "%= %3n %t%? [%h]%? %="
#caption always "%= %-w%L>%{= BW}%n*%t%{-}%52<%+w %L="


hardstatus alwayslastline
hardstatus string '%{= kG}[%{G}%H%? %1`%?%{g}][%= %{= kw}%-w%{+b yk} %n*%t%?(%u)%? %{-}%+w %=%{g}][%{B}%m/%d %{W}%C%A%{g}]'

defscrollback 3000

jelsas

#!/usr/bin/python
'''
A command-line calculator over streams of numbers.
Requires numpy for 'hist' and 'median'.

  For example:

# create a file with 100 random numbers, 1 per line
$ jot -r 100 > /tmp/random

jelsas

'''
Functions to build full dependence model queries for Indri http://lemurproject.org/indri.
See "A Markov Random Field Model for Term Dependencies" by Metzler & Croft
http://ciir.cs.umass.edu/pubfiles/ir-387.pdf
'''
import re

nonword_chars = re.compile(r'\W+')

def powerset(l):

jelsas

#!/usr/bin/python

# This script takes the predictions and input vectors
# (eg. set1.test.txt) and produces a file suitable for
# submission to the Yahoo LETOR Challenge web interface.
#
# The PREDICTIONS_FILE should just be a list of scores,
# one per line, corresponding to the lines in INPUT_FILE

from itertools import izip, groupby

jelsas

% This is a drop-in replacement for \begin{enumerate}...\end{enumerate} in LaTeX.  Just include it in
% your preamble and use \begin{packed_enum} instead.
\newenvironment{packed_enum}{
\begin{enumerate}
  \setlength{\itemsep}{2.5pt}
  \setlength{\parskip}{0pt}
  \setlength{\parsep}{0pt}
}{\end{enumerate}}