def run_rep_system():
# Gets the information on the venue.
c = db.venue(request.args[0]) or redirect(URL('default', 'index'))
check_manager_eligibility(c.id, auth.user.email, 'You cannot evaluate contributors for this venue')
# This venue_form is used to display the venue.
venue_form = SQLFORM(db.venue, record=c, readonly=True)
confirmation_form = FORM.confirm(T('Run'),
{T('Cancel'): URL('venues', 'view_venue', args=[c.id])})
if confirmation_form.accepted:
num_of_iterations = 4
ranker.run_reputation_system(db, c.id,
num_of_iterations=num_of_iterations)
db.commit()
session.flash = T('The computation of reviewer contribution, submission quality, and final grade is complete.')
redirect(URL('venues', 'view_venue', args=[c.id]))
return dict(venue_form=venue_form, confirmation_form=confirmation_form)
def run_rep_system():
# Gets the information on the venue.
c = db.venue(request.args(0)) or redirect(URL('default', 'index'))
check_manager_eligibility(
c.id, auth.user.email,
'You cannot evaluate contributors for this venue')
# This venue_form is used to display the venue.
venue_form = SQLFORM(db.venue, record=c, readonly=True)
confirmation_form = FORM.confirm(
T('Run'), {T('Cancel'): URL('venues', 'view_venue', args=[c.id])})
if confirmation_form.accepted:
num_of_iterations = 4
ranker.run_reputation_system(db,
c.id,
num_of_iterations=num_of_iterations)
db.commit()
session.flash = T(
'The computation of reviewer contribution, submission quality, and final grade is complete.'
)
redirect(URL('venues', 'view_venue', args=[c.id]))
return dict(venue_form=venue_form, confirmation_form=confirmation_form)
def run_rep_sys_research():
# Gets the information on the venue.
c = db.venue(request.args(0)) or redirect(URL('default', 'index'))
rep_sys_type = int(request.args(1)) or redirect(URL('default', 'index'))
check_manager_eligibility(c.id, auth.user.email, 'Not authorized.')
# This venue_form is used to display the venue.
venue_form = SQLFORM(db.venue, record=c, readonly=True)
confirmation_form = FORM.confirm(
T('Run'), {T('Cancel'): URL('venues', 'view_venue', args=[c.id])})
if confirmation_form.accepted:
if rep_sys_type == 1:
# Run without reputation system
ranker.rank_without_rep_sys(db, c.id, alpha_annealing=0.5)
elif rep_sys_type == 2:
# Run reputation system on all comparisons in chronological
# order one time.
ranker.run_reputation_system(db,
c.id,
alpha_annealing=0.5,
last_compar_param=None)
elif rep_sys_type == 3:
# Run reputation system with small alpha on latest comparisons.
ranker.run_reputation_system(db, c.id, num_of_iterations=4)
else:
# Run the latest reputation system.
ranker.run_reputation_system(db, c.id, num_of_iterations=4)
db.commit()
session.flash = T(
'The computation of reviewer contribution, submission quality, and final grade is complete.'
)
redirect(URL('venues', 'view_venue_research', args=[c.id]))
# Description of reputation system.
if rep_sys_type == 1:
description = T(
"Ranking without reputation system. All comparisons are used in chronological order."
)
elif rep_sys_type == 2:
description = T(
"Reputation system on all comparisons in chronological order.")
elif rep_sys_type == 3:
description = T(
"Reputation system with small alpha and only last comparisons.")
else:
description = T(
"Reputation system with small alpha and only last comparisons.")
return dict(venue_form=venue_form,
confirmation_form=confirmation_form,
rep_sys_description=description)
def run_rep_sys_research():
# Gets the information on the venue.
c = db.venue(request.args(0)) or redirect(URL('default', 'index'))
rep_sys_type = int(request.args(1)) or redirect(URL('default', 'index'))
check_manager_eligibility(c.id, auth.user.email, 'Not authorized.')
# This venue_form is used to display the venue.
venue_form = SQLFORM(db.venue, record=c, readonly=True)
confirmation_form = FORM.confirm(T('Run'),
{T('Cancel'): URL('venues', 'view_venue', args=[c.id])})
if confirmation_form.accepted:
if rep_sys_type == 1:
# Run without reputation system
ranker.rank_without_rep_sys(db, c.id, alpha_annealing=0.5)
elif rep_sys_type == 2:
# Run reputation system on all comparisons in chronological
# order one time.
ranker.run_reputation_system(db, c.id, alpha_annealing=0.5,
last_compar_param=None)
elif rep_sys_type == 3:
# Run reputation system with small alpha on latest comparisons.
ranker.run_reputation_system(db, c.id, num_of_iterations=4)
else:
# Run the latest reputation system.
ranker.run_reputation_system(db, c.id, num_of_iterations=4)
db.commit()
session.flash = T('The computation of reviewer contribution, submission quality, and final grade is complete.')
redirect(URL('venues', 'view_venue_research', args=[c.id]))
# Description of reputation system.
if rep_sys_type == 1:
description = T("Ranking without reputation system. All comparisons are used in chronological order.")
elif rep_sys_type == 2:
description = T("Reputation system on all comparisons in chronological order.")
elif rep_sys_type == 3:
description = T("Reputation system with small alpha and only last comparisons.")
else:
description = T("Reputation system with small alpha and only last comparisons.")
return dict(venue_form=venue_form, confirmation_form=confirmation_form,
rep_sys_description=description)
def run_rep_sys():
"""Runs the reputation system."""
c = db.venue(request.vars[REPUTATION_SYSTEM_PARAM_VENUE_ID])
if c is None:
logger.warning("Invalid call to reputation system computation: " + str(request))
return redirect(URL('venues', 'managed_index'))
try:
current.num_iterations = int(request.vars[REPUTATION_SYSTEM_PARAM_NUM_ITERATIONS])
except Exception:
logger.warning("Missing number of iterations in specification.")
current.num_iterations = ALGO_DEFAULT_NUM_ITERATIONS
try:
current.review_percentage = float(request.vars[REPUTATION_SYSTEM_PARAM_REVIEW_PERCENTAGE])
except Exception:
current.review_percentage = ALGO_DEFAULT_REVIEWS_AS_PERCENTAGE
startover = False
try:
startover = request.vars[REPUTATION_SYSTEM_STARTOVER] == 'True'
except Exception:
pass
run_id = request.vars[REPUTATION_SYSTEM_RUN_ID]
publish = request.vars[REPUTATION_SYSTEM_PUBLISH] == 'True'
algo = request.vars[REPUTATION_SYSTEM_ALGO]
current.cost_type = request.vars[REPUTATION_SYSTEM_COST_TYPE]
if current.cost_type is None:
current.cost_type = ALGO_DEFAULT_COST_TYPE
try:
current.pos_slope = float(request.vars[REPUTATION_SYSTEM_POS_SLOPE])
except Exception:
current.pos_slope = ALGO_DEFAULT_POS_SLOPE
try:
current.neg_slope = float(request.vars[REPUTATION_SYSTEM_NEG_SLOPE])
except Exception:
current.neg_slope = ALGO_DEFAULT_NEG_SLOPE
current.normalize_grades = request.vars[REPUTATION_SYSTEM_NORMALIZE_GRADES] == 'True'
try:
current.normalization_scale = float(request.vars[REPUTATION_SYSTEM_NORMALIZATION_SCALE])
except Exception:
current.normalization_scale = ALGO_DEFAULT_NORMALIZATION_SCALE
current.use_submission_rank_in_rep = request.vars[REPUTATION_SYSTEM_USE_SUBMISSION_RANK_IN_REP] == 'True'
try:
current.submission_rank_exp = float(request.vars[REPUTATION_SYSTEM_SUBMISSION_RANK_REP_EXP])
except Exception:
current.submission_rank_exp = ALGO_DEFAULT_RANK_REP_EXP
current.reputation_method = request.vars[REPUTATION_SYSTEM_REPUTATION_METHOD]
if current.reputation_method is None:
current.reputation_method = ALGO_DEFAULT_REPUTATION_METHOD
try:
current.prec_coefficient = float(request.vars[REPUTATION_SYSTEM_PREC_COEFF])
except Exception:
current.prec_coefficient = ALGO_DEFAULT_PREC_COEFF
try:
current.matrix_D_type = request.vars[REPUTATION_SYSTEM_MATRIX_D_TYPE]
except Exception:
current.matrix_D_type = MATRIX_D_TYPE_GRADES_DIST
current.precision_method = request.vars[REPUTATION_SYSTEM_PREC_METHOD]
if current.precision_method is None:
current.precision_method = ALGO_DEFAULT_PREC_METHOD
# Vancouver.
current.vancouver_use_median = request.vars[REPUTATION_SYSTEM_USE_MEDIAN] != 'False'
current.vancouver_do_debias = request.vars[REPUTATION_SYSTEM_DO_DEBIAS] == 'True'
current.vancouver_use_reputation = request.vars[REPUTATION_SYSTEM_USE_REPUTATION] != 'False'
logger.info("Reputation system request: %r" % request.vars)
logger.info("Starting reputation system run for venue: " + c.name)
logger.info("Requested number of iterations: %d" % current.num_iterations)
logger.info("Review percentage: %f" % current.review_percentage)
logger.info("Using algo: %r" % algo)
logger.info("Run id: %r" % run_id)
logger.info("Publish: %s" % publish)
logger.info("Number of iterations: %s" % current.num_iterations)
logger.info("Pos slope: %r" % current.pos_slope)
logger.info("Neg slope: %r" % current.neg_slope)
logger.info("Normalize grades: %r" % current.normalize_grades)
logger.info("Normalization scale: %r" % current.normalization_scale)
logger.info("Use submission rank in rep: %r" % current.use_submission_rank_in_rep)
logger.info("Submission rank exponent for reputation: %r" % current.submission_rank_exp)
logger.info("Reputation method: %r" % current.reputation_method)
logger.info("Vancouver use median: %r" % current.vancouver_use_median)
logger.info("Vancouver do debias: %r" % current.vancouver_do_debias)
logger.info("Vancouver use reputation: %r" % current.vancouver_use_reputation)
logger.info("Precision coefficient: %r" % current.prec_coefficient)
logger.info("Type of matrix D: %r" % current.matrix_D_type)
# Stores the run parameters.
db.run_parameters.update_or_insert(
(db.run_parameters.venue_id == c.id) & (db.run_parameters.run_id == run_id),
venue_id = c.id,
run_id = run_id,
params = json.dumps(request.vars))
if algo == ALGO_OPT:
grades_rank.rank_by_grades(c.id, run_id=run_id, publish=publish)
elif algo == ALGO_VANCOUVER:
v = vancouver.Vancouver(c.id, run_id=run_id, publish=publish,
use_median=current.vancouver_use_median,
do_debias=current.vancouver_do_debias)
v.run_evaluation(use_reputation=current.vancouver_use_reputation)
else:
# DEPRECATED
ranker.run_reputation_system(c.id, num_of_iterations=current.num_iterations,
review_percentage=current.review_percentage,
startover=startover, publish=publish,
run_id=run_id)
logger.info("Completed reputation system run for venue: " + c.name)
return redirect(URL('ranking', 'view_grades', args=[c.id]))
def run_rep_sys():
"""Runs the reputation system."""
c = db.venue(request.vars[REPUTATION_SYSTEM_PARAM_VENUE_ID])
if c is None:
logger.warning("Invalid call to reputation system computation: " +
str(request))
return redirect(URL('venues', 'managed_index'))
try:
current.num_iterations = int(
request.vars[REPUTATION_SYSTEM_PARAM_NUM_ITERATIONS])
except Exception:
logger.warning("Missing number of iterations in specification.")
current.num_iterations = ALGO_DEFAULT_NUM_ITERATIONS
try:
current.review_percentage = float(
request.vars[REPUTATION_SYSTEM_PARAM_REVIEW_PERCENTAGE])
except Exception:
current.review_percentage = ALGO_DEFAULT_REVIEWS_AS_PERCENTAGE
startover = False
try:
startover = request.vars[REPUTATION_SYSTEM_STARTOVER] == 'True'
except Exception:
pass
run_id = request.vars[REPUTATION_SYSTEM_RUN_ID]
publish = request.vars[REPUTATION_SYSTEM_PUBLISH] == 'True'
algo = request.vars[REPUTATION_SYSTEM_ALGO]
current.cost_type = request.vars[REPUTATION_SYSTEM_COST_TYPE]
if current.cost_type is None:
current.cost_type = ALGO_DEFAULT_COST_TYPE
try:
current.pos_slope = float(request.vars[REPUTATION_SYSTEM_POS_SLOPE])
except Exception:
current.pos_slope = ALGO_DEFAULT_POS_SLOPE
try:
current.neg_slope = float(request.vars[REPUTATION_SYSTEM_NEG_SLOPE])
except Exception:
current.neg_slope = ALGO_DEFAULT_NEG_SLOPE
current.normalize_grades = request.vars[
REPUTATION_SYSTEM_NORMALIZE_GRADES] == 'True'
try:
current.normalization_scale = float(
request.vars[REPUTATION_SYSTEM_NORMALIZATION_SCALE])
except Exception:
current.normalization_scale = ALGO_DEFAULT_NORMALIZATION_SCALE
current.use_submission_rank_in_rep = request.vars[
REPUTATION_SYSTEM_USE_SUBMISSION_RANK_IN_REP] == 'True'
try:
current.submission_rank_exp = float(
request.vars[REPUTATION_SYSTEM_SUBMISSION_RANK_REP_EXP])
except Exception:
current.submission_rank_exp = ALGO_DEFAULT_RANK_REP_EXP
current.reputation_method = request.vars[
REPUTATION_SYSTEM_REPUTATION_METHOD]
if current.reputation_method is None:
current.reputation_method = ALGO_DEFAULT_REPUTATION_METHOD
try:
current.prec_coefficient = float(
request.vars[REPUTATION_SYSTEM_PREC_COEFF])
except Exception:
current.prec_coefficient = ALGO_DEFAULT_PREC_COEFF
try:
current.matrix_D_type = request.vars[REPUTATION_SYSTEM_MATRIX_D_TYPE]
except Exception:
current.matrix_D_type = MATRIX_D_TYPE_GRADES_DIST
current.precision_method = request.vars[REPUTATION_SYSTEM_PREC_METHOD]
if current.precision_method is None:
current.precision_method = ALGO_DEFAULT_PREC_METHOD
# Vancouver.
current.vancouver_use_median = request.vars[
REPUTATION_SYSTEM_USE_MEDIAN] != 'False'
current.vancouver_do_debias = request.vars[
REPUTATION_SYSTEM_DO_DEBIAS] == 'True'
current.vancouver_use_reputation = request.vars[
REPUTATION_SYSTEM_USE_REPUTATION] != 'False'
logger.info("Reputation system request: %r" % request.vars)
logger.info("Starting reputation system run for venue: " + c.name)
logger.info("Requested number of iterations: %d" % current.num_iterations)
logger.info("Review percentage: %f" % current.review_percentage)
logger.info("Using algo: %r" % algo)
logger.info("Run id: %r" % run_id)
logger.info("Publish: %s" % publish)
logger.info("Number of iterations: %s" % current.num_iterations)
logger.info("Pos slope: %r" % current.pos_slope)
logger.info("Neg slope: %r" % current.neg_slope)
logger.info("Normalize grades: %r" % current.normalize_grades)
logger.info("Normalization scale: %r" % current.normalization_scale)
logger.info("Use submission rank in rep: %r" %
current.use_submission_rank_in_rep)
logger.info("Submission rank exponent for reputation: %r" %
current.submission_rank_exp)
logger.info("Reputation method: %r" % current.reputation_method)
logger.info("Vancouver use median: %r" % current.vancouver_use_median)
logger.info("Vancouver do debias: %r" % current.vancouver_do_debias)
logger.info("Vancouver use reputation: %r" %
current.vancouver_use_reputation)
logger.info("Precision coefficient: %r" % current.prec_coefficient)
logger.info("Type of matrix D: %r" % current.matrix_D_type)
# Stores the run parameters.
db.run_parameters.update_or_insert((db.run_parameters.venue_id == c.id) &
(db.run_parameters.run_id == run_id),
venue_id=c.id,
run_id=run_id,
params=json.dumps(request.vars))
if algo == ALGO_OPT:
grades_rank.rank_by_grades(c.id, run_id=run_id, publish=publish)
elif algo == ALGO_VANCOUVER:
v = vancouver.Vancouver(c.id,
run_id=run_id,
publish=publish,
use_median=current.vancouver_use_median,
do_debias=current.vancouver_do_debias)
v.run_evaluation(use_reputation=current.vancouver_use_reputation)
else:
# DEPRECATED
ranker.run_reputation_system(
c.id,
num_of_iterations=current.num_iterations,
review_percentage=current.review_percentage,
startover=startover,
publish=publish,
run_id=run_id)
logger.info("Completed reputation system run for venue: " + c.name)
return redirect(URL('ranking', 'view_grades', args=[c.id]))
