Set

In jajapy, a training / test set is a member of the Set class described below.

A Set has three attributes:

sequences: which is the list of all possible traces in the set.
times: which corresponds to the number of times each traces appears in the set.
type: that indicates which kind of traces are in sequences.

>>> import jajapy as ja
>>> model = ja.HMM_random(3,['a','b'])
>>> s = model.generateSet(5,3)
>>> s.sequences
[['a','b','b'],['a','b','a'],['b','a','a']]
>>> s.times
[1,3,1]
>>> # model has generated 'abb' and 'baa' once, and 'aba' three times
>>> s.type
0
>>> # if s.type == 0 then the traces are sequences of discrete observations,
>>> # i.e. they have been generated by a HMM or a MC (they can also be non-timed
>>> # traces of a CTMC).

class jajapy.Set(sequences: list, times: Optional[list] = None, t: Optional[int] = None)

Class representing a set (training set / test set).

Creates the Set.

Parameters

sequenceslist

List of sequences of traces. A trace can be: - a sequence of observations (list of str) ->HMM or MC - an alternating sequence of actions/observations (list of str) ->MDP - a sequence of vectors of timed observations (list of list of float) ->GoHMM - an alternating sequence of waiting-times/observations (list of float and str) ->CTMC

timeslist, optional

times[i] is the number of time sequences[i] appears in the set. Can be omitted.

tint, optional

0 if this set was generated by a HMM or a MC,
1 if it was generated by a MDP,
3 if it was generated by a GoHMM,
4 if it was generated by a CTMC.

addSet(s2)

Merges this set with another generated by the same kind of Markov model.

Parameters

s2Set: set 2.

Raises

TypeError: If the two models type attribute are different.

getActionsObservations() → list

Returns all possible observations and all possible actions in this set.

Returns

list: list of one list of actions and one list of observations.

getAlphabet() → list

Returns the list of all possible observations in this set.

Returns

list: list of observations.

isEqual(set2) → bool

Checks wheter or not this set is equal to another set set2.

Parameters

set2Set: Another set.

Returns

bool: True if this set is equal to set2.

save(file_path: str) → None

Save this set into a text file.

Parameters

file_path: str: where to save

setFromList(l: list) → None

Convert a list of sequences of observations to a set.

Parameters

llist: list of sequences of observations.

sort() → None: Sort the sequences in lexicographic order.