def __init__(self, DEBUG, list_of_demonstrations, featfile, trialname):
# self.list_of_demonstrations = parser.generate_list_of_videos(constants.PATH_TO_DATA + constants.CONFIG_FILE)
self.list_of_demonstrations = list_of_demonstrations
self.data_X = {}
self.data_X_size = {}
self.data_N = {}
self.file = None
self.featfile = featfile
self.metrics_picklefile = None
self.change_pts = None
self.change_pts_Z = None
self.change_pts_W = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp_level1 = {}
self.map_cp_level2 = {}
self.map_level1_cp = {}
self.map_level2_cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.l2_cluster_matrices = {}
self.map_frm2surgeme = parser.get_all_frame2surgeme_maps(self.list_of_demonstrations)
self.trial = utils.hashcode() + trialname
# self.trial = trialname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.pruned_L2_clusters = []
self.silhouette_scores = {}
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.level2_dunn_1 = []
self.level2_dunn_2 = []
self.level2_dunn_3 = []
self.level2_silhoutte = []
self.label_based_scores_1 = {}
self.label_based_scores_2 = {}
self.gmm_objects = {}
self.sr = 10
def __init__(self, DEBUG, list_of_demonstrations, featfile, trialname):
# self.list_of_demonstrations = parser.generate_list_of_videos(constants.PATH_TO_DATA + constants.CONFIG_FILE)
self.list_of_demonstrations = list_of_demonstrations
self.data_X = {}
self.data_X_size = {}
self.data_N = {}
self.file = None
self.featfile = featfile
self.metrics_picklefile = None
self.change_pts = None
self.change_pts_Z = None
self.change_pts_W = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp_level1 = {}
self.map_cp_level2 = {}
self.map_level1_cp = {}
self.map_level2_cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.l2_cluster_matrices = {}
self.map_frm2surgeme = utils.get_all_frame2surgeme_maps(self.list_of_demonstrations)
self.trial = utils.hashcode() + trialname
# self.trial = trialname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.pruned_L2_clusters = []
self.silhouette_scores = {}
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.level2_dunn_1 = []
self.level2_dunn_2 = []
self.level2_dunn_3 = []
self.level2_silhoutte = []
self.label_based_scores_1 = {}
self.label_based_scores_2 = {}
self.gmm_objects = {}
self.sr = 10
from save.instituicoes_pt import *
serialize_rdf_unidades = {
"classType":
Unidade,
"collection": [
{ ## ufrn
"toSave":
False,
"mapper": {
"nome":
"nome_unidade",
"code":
"id_unidade",
"id":
lambda d: hashcode("ufrn", "centro", d["id_unidade"]),
"university":
lambda d: UFRN,
"sameas":
lambda d:
"https://sigaa.ufrn.br/sigaa/public/departamento/portal.jsf?id="
+ str(d["id_unidade"]),
},
"data":
lambda: list(
filter(
lambda d: d["tipo_unidade_organizacional"].find("CENTRO") >
-1,
dados_ckan(
"http://dados.ufrn.br/api/action/datastore_search?resource_id=3f2e4e32-ef1a-4396-8037-cbc22a89d97f"
))),
def __init__(self, DEBUG, list_of_demonstrations, featfile, trialname):
self.list_of_demonstrations = list_of_demonstrations
self.data_X = {}
self.data_W = {}
self.data_Z = {}
self.data_X_size = {}
self.data_N = {}
self.file = None
self.featfile = featfile
self.metrics_picklefile = None
self.change_pts = None
self.change_pts_Z = None
self.change_pts_W = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp2level1 = {}
self.map_level12cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.l2_cluster_matrices = {}
self.map_frm2surgeme = utils.get_all_frame2surgeme_maps(self.list_of_demonstrations)
self.trial = utils.hashcode() + trialname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.pruned_L2_clusters = []
self.silhouette_scores_global = {}
self.silhouette_scores_weighted = {}
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.level2_dunn_1 = []
self.level2_dunn_2 = []
self.level2_dunn_3 = []
self.level2_silhoutte_global = []
self.level2_silhoutte_weighted = []
self.label_based_scores_1 = {}
self.label_based_scores_2 = {}
self.sr = constants.SR
self.representativeness = constants.PRUNING_FACTOR_ZW
# Components for Mixture model at each level
self.n_components_cp = constants.N_COMPONENTS_CP
self.n_components_L1 = constants.N_COMPONENTS_L1
self.n_components_L2 = constants.N_COMPONENTS_L2
self.temporal_window = constants.TEMPORAL_WINDOW_ZW
self.fit_GMM = False
self.fit_DPGMM = True
assert self.fit_DPGMM or self.fit_GMM == True
from utils import dados_ckan, dados_ufma, hashcode, dados_iffar
from save.instituicoes_pt import *
serialize_rdf_subunidades = {
"classType":
Subunidade,
"collection": [
{ ## ufrn
"toSave":
False,
"mapper": {
"nome":
"nome_unidade",
"id":
lambda d: hashcode("ufrn", "departamento", d["id_unidade"]),
"code":
"id_unidade",
"university":
lambda d: UFRN,
"unidade":
lambda d: "https://www.dbacademic.tech/resource/" + hashcode(
"ufrn", "centro", d["id_unidade_responsavel"]),
"sameas":
lambda d:
"https://sigaa.ufrn.br/sigaa/public/departamento/portal.jsf?id="
+ str(d["id_unidade"]),
},
"data":
lambda: list(
filter(
def __init__(self, DEBUG, list_of_demonstrations, featfile, trialname):
self.list_of_demonstrations = list_of_demonstrations
self.data_X = {}
self.data_W = {}
self.data_Z = {}
self.data_X_size = {}
self.data_N = {}
self.file = None
self.featfile = featfile
self.metrics_picklefile = None
self.change_pts = None
self.change_pts_Z = None
self.change_pts_W = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp2level1 = {}
self.map_level12cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.l2_cluster_matrices = {}
self.map_frm2surgeme = utils.get_all_frame2surgeme_maps(
self.list_of_demonstrations)
self.trial = utils.hashcode() + trialname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.pruned_L2_clusters = []
self.silhouette_scores_global = {}
self.silhouette_scores_weighted = {}
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.level2_dunn_1 = []
self.level2_dunn_2 = []
self.level2_dunn_3 = []
self.level2_silhoutte_global = []
self.level2_silhoutte_weighted = []
self.label_based_scores_1 = {}
self.label_based_scores_2 = {}
self.sr = constants.SR
self.representativeness = constants.PRUNING_FACTOR_ZW
# Components for Mixture model at each level
self.n_components_cp = constants.N_COMPONENTS_CP
self.n_components_L1 = constants.N_COMPONENTS_L1
self.n_components_L2 = constants.N_COMPONENTS_L2
self.temporal_window = constants.TEMPORAL_WINDOW_ZW
self.fit_GMM = False
self.fit_DPGMM = True
assert self.fit_DPGMM or self.fit_GMM == True
def __init__(self, DEBUG, list_of_demonstrations, fname, log, vision_mode = False, feat_fname = None):
self.list_of_demonstrations = list_of_demonstrations
if vision_mode and feat_fname is None:
print "[Error] Please provide file with visual features"
sys.exit()
self.vision_mode = vision_mode
self.feat_fname = feat_fname
self.data_X = {}
self.X_dimension = 0
self.data_X_size = 0
self.data_N = {}
self.log = log
self.file = None
self.metrics_picklefile = constants.PATH_TO_CLUSTERING_RESULTS + fname + ".p"
self.changepoints = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp2cluster = {}
self.map_level1_cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.map_frm2surgeme = utils.get_all_frame2surgeme_maps(self.list_of_demonstrations)
self.trial = utils.hashcode() + fname
# self.trial = fname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.silhouette_score_global = None
self.silhouette_score_weighted = None
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.label_based_scores_1 = {}
self.sr = constants.SR
# Components for Mixture model at each level
if self.vision_mode:
self.n_components_cp = constants.N_COMPONENTS_CP_Z
self.n_components_L1 = constants.N_COMPONENTS_L1_Z
self.temporal_window = constants.TEMPORAL_WINDOW_Z
self.representativeness = constants.PRUNING_FACTOR_Z
self.ALPHA_CP = constants.ALPHA_Z_CP
if (constants.TASK_NAME in ["000", "010", "011", "100"]):
self.ALPHA_CP = constants.ALPHA_W_CP
else:
self.n_components_cp = constants.N_COMPONENTS_CP_W
self.n_components_L1 = constants.N_COMPONENTS_L1_W
self.temporal_window = constants.TEMPORAL_WINDOW_W
self.representativeness = constants.PRUNING_FACTOR_W
self.ALPHA_CP = constants.ALPHA_W_CP
self.ALPHA_L1 = 0.4
self.fit_GMM = False
self.fit_DPGMM = True
assert self.fit_DPGMM or self.fit_GMM == True
def __init__(self, DEBUG, list_of_demonstrations, fname, log, vision_mode = False, feat_fname = None):
self.list_of_demonstrations = list_of_demonstrations
if vision_mode and feat_fname is None:
print "[Error] Please provide file with visual features"
sys.exit()
self.vision_mode = vision_mode
self.feat_fname = feat_fname
self.data_X = {}
self.X_dimension = 0
self.data_X_size = 0
self.data_N = {}
self.log = log
self.file = None
self.metrics_picklefile = constants.PATH_TO_CLUSTERING_RESULTS + fname + ".p"
self.changepoints = None
self.list_of_cp = []
self.map_cp2frm = {}
self.map_cp2demonstrations = {}
self.map_cp2cluster = {}
self.map_level1_cp = {}
self.map_cp2milestones = {}
self.map_cp2surgemes = {}
self.map_cp2surgemetransitions = {}
self.map_frm2surgeme = parser.get_all_frame2surgeme_maps(self.list_of_demonstrations)
self.trial = utils.hashcode() + fname
# self.trial = fname
self.cp_surgemes = []
self.pruned_L1_clusters = []
self.silhouette_score_global = None
self.silhouette_score_weighted = None
self.dunn_scores_1 = {}
self.dunn_scores_2 = {}
self.dunn_scores_3 = {}
self.label_based_scores_1 = {}
self.sr = constants.SR
# Components for Mixture model at each level
if self.vision_mode:
self.n_components_cp = constants.N_COMPONENTS_CP_Z
self.n_components_L1 = constants.N_COMPONENTS_L1_Z
self.temporal_window = constants.TEMPORAL_WINDOW_Z
self.representativeness = constants.PRUNING_FACTOR_Z
self.ALPHA_CP = constants.ALPHA_Z_CP
if (constants.TASK_NAME in ["000", "010", "011", "100"]):
self.ALPHA_CP = constants.ALPHA_W_CP
else:
self.n_components_cp = constants.N_COMPONENTS_CP_W
self.n_components_L1 = constants.N_COMPONENTS_L1_W
self.temporal_window = constants.TEMPORAL_WINDOW_W
self.representativeness = constants.PRUNING_FACTOR_W
self.ALPHA_CP = constants.ALPHA_W_CP
self.ALPHA_L1 = 0.4
self.fit_GMM = False
self.fit_DPGMM = True
assert self.fit_DPGMM or self.fit_GMM == True
from utils import dados_ckan, dados_ufma, hashcode, removeNonUTF8, dados_csv
from save.instituicoes_pt import *
serialize_rdf_monografia = {
"classType":
Monografia,
"collection": [
{ ## ufrn
"toSave":
False,
"mapper": {
"titulo": lambda d: removeNonUTF8(d["titulo"]),
"university": lambda d: UFRN,
"id": lambda d: hashcode("ufrn", "monografias", str(d["_id"])),
"autor": "nome_autor"
},
"data":
lambda: dados_ckan(
"http://dados.ufrn.br/api/action/datastore_search?resource_id=7c01071b-81a4-4793-9a63-acfcd8a1aa83"
),
"rdf_path":
"rdf/monografias_ufrn.rdf"
},
{ ## ufma
"toSave":
False,
"mapper": {
"titulo":
lambda d: (d["titulo"]),
