| | |
Methods defined here:
- __init__(self, data_package, wilc_data_type='probe', weight=None, scale=None, minWeight=0.5)
- This class generates a sequence of learner objects.
Inputs:
req:
data_package: provide an already set up packager.DataPackage object
optional:
wilc_data_type: should we use probes or genes to simulate the wilcoxon process in the TSx's
weight: A list of weights that estimated times are divided by
scale: A list of scaling factors for the complexity estimates.
minWeight: a weight at which to normalize the weights.
- __iter__(self)
- feedback(self, learner_id, apparent_accuracy)
- Adjusts the weights of a learner.
learner_id matching global data attribute
apparent_accuracy - percent accuracy of learner
- genDirac(self, min_network_size, numTopNetworks, data_type='gene')
- generates the pq for dirac.
inputs:
min_network_size: a tuple with (start, end, increment)
numTopNetworks: a tuple with (start, end, increment)
- genKTSP(self, maxK, ncv, nlo, r1, r2, equijoin=False, data_type='probe')
- generates the pq for tst
inputs:
maxK - tuple describing the range for the maximum k value
ncv - tuple describing the range for number of cross validations
nlo - tuple describing the range for number of elements to leave out of internal crossvalidation
r1 - tuple describing the range for filter 1 (from, to, increment)
r2 - tuple describing the range for filter 2 (from, to, increment)
equijoin - boolean, should we restrict filters to [10,10] [20,20] etc.
- genTSP(self, r1, r2, equijoin=False, data_type='probe')
- generates the pq for tsp.
inputs:
r1 - tuple describing the range for filter 1 (from, to, increment)
r2 - tuple describing the range for filter 2 (from, to, increment)
equijoin - boolean, should we restrict filters to [10,10] [20,20] etc.
- genTST(self, r1, r2, r3, equijoin=False, data_type='probe')
- generates the pq for tst
inputs:
r1 - tuple describing the range for filter 1 (from, to, increment)
r2 - tuple describing the range for filter 2 (from, to, increment)
r3 - tuple describing the range for filter 3 (from, to, increment)
equijoin - boolean, should we restrict filters to [10,10] [20,20] etc.
- getEstimatedTime(self, learner_id, complexity)
- getLearner(self, settings)
- Returns a learner object corresponding to the provided settings dict.
- next(self)
- trainLearner(self, settings, complexity)
- Returns a learner that has been trained according to the settings given.
Data and other attributes defined here:
- dirac = 0
- ktsp = 3
- tsp = 1
- tst = 2
|