#!/usr/bin/env python
'''
cut.py -- cut specified columns from a TSV-formatted file.
Assumes the file contains a header line containing column names,
of which the columns to be cut are subset.
Modify the columns variable below.
'''
import sys
from datastore.table import Reader
transcripts = sys.argv[1:]
columns = '''id subject session row time line key
p_utts_orig p_utts p_form p_lrb p_obj p_gloss p_orient p_mspd
c_utts_orig c_utts c_form c_lrb c_obj c_gloss c_orient c_mspd
context'''.split()
def pprint(values): print "\t".join(values)
pprint(columns)
for t in transcripts:
T = Reader(t)
for row in T.values(*columns):
pprint(row)
(19,86), (20,90)])
speech = Reader('speech.xls')
ses = Reader('ses.xls')
subjects = defaultdict(dict)
ses_cols = 'SUBJ SEX EDU INC RACE ETHN'.split()
for row in ses:
subjects[row['SUBJ']] = row
visits = defaultdict(dict)
columns = 'subject session speaker word_types'.split()
for subj, sess, spkr, wt in speech.values(*columns):
age = sess_map[int(sess)]
if not visits.has_key((subj, sess)):
visits[subj, sess] = {'SUBJ': subj, 'SESS': sess, 'AGE': age,
'CWT': '', 'PWT': ''}
if spkr == "child":
visits[subj, sess]['CWT'] = wt
else:
visits[subj, sess]['PWT'] = wt
viz_cols = 'SESS AGE PWT CWT'.split()
print "\t".join(ses_cols + viz_cols)
for id, data in visits.items():
subj, sess = id
from collections import defaultdict
ses = Reader('ses.xls')
subjects = defaultdict(dict)
ses_cols = 'SUBJ SEX EDU INC RACE ETHN'.split()
for row in ses:
subjects[row['SUBJ']] = row
outcomes = Reader('outcomes.tsv')
out = defaultdict(dict)
columns = 'SUBJ SESS VOCAB READ_WJ READ_GM'.split()
for subj, sess, voc, rwj, rgm in outcomes.values(*columns):
if not out.has_key(subj):
out[subj] = {
'SUBJ': subj,
'VOCB1': '',
'VOCB2': '',
'VOCB3': '',
'VOCB4': '',
'READ1': '',
'READ2': '',
'READ3': '',
'READ4': '',
'READ5': ''
}
if sess == "5" and voc:
out[subj]['VOCB1'] = voc
from datastore.table import Reader
r = Reader('ses.tsv')
def pprint(args):
print "\t".join(args)
pprint('SUBJ SEX EDU INC RACE ETHN'.split())
for v in r.values('id', 'sex', 'edu', 'income', 'race', 'ethn'):
pprint(v)