def __init__(self, pd, transform=None):
"""
Args:
pd: Pandas dataframe, first column is assumed to be data
transform (callable, optional): Optional transform to be applied
on a sample.
"""
self.pd_frame = pd.copy()
self.transform = transform
def binary_n(pd):
rs = pd.copy()
for row in pd.iterrows():
row = row[1]
rows = binary(row)
if rows is None:
continue
rs = rs.append(rows)
print len(rs)
print len(pd)
return rs
def alg():
start = 0
koniec = start
rozmiar_macierzy_wag = 30
rozmiar_populacji = 30
liczba_iteracji = 1000
liczba_par = 9
liczba_najlepszych_rozwiazan = 2
prawdopodobienstwo_mutacji = 0.05
liczba_grup = 1
macierz = macierz_wag(rozmiar_macierzy_wag)
populacja = populacja_startowa(rozmiar_populacji, macierz, start, koniec)
ostatnia_odl = 10000000
najlepsze_wyniki = []
for i in range(liczba_iteracji):
nowa_populacja = []
# drogi bieżącej populacji
wyniki = wynik_populacji(populacja, macierz)
najlepszy = populacja[np.argmin(wyniki)] # najkrotsza droga
liczba_ruchow = len(najlepszy)
odleglosc = zmierz_dlugosc(najlepszy, macierz)
if odleglosc != ostatnia_odl:
print('Iteracja %i: droga wynosi %f' % (i, odleglosc))
# rozmnażanie się członków na podstawanie wyników podobieństwa/metoda ruletki
for j in range(liczba_par):
nowy_1 = krzyzowanie(list(populacja[znajdz_kolege(wyniki)]),
list(populacja[znajdz_kolege(wyniki)]))
nowa_populacja = nowa_populacja + [nowy_1]
# mutacja
for j in range(len(nowa_populacja)):
nowa_populacja[j] = np.copy(
mutacja(nowa_populacja[j], prawdopodobienstwo_mutacji,
macierz))
# zatrzymanie członków starej populacjiw w nowej
nowa_populacja += [populacja[np.argmin(wyniki)]]
for j in range(1, liczba_najlepszych_rozwiazan):
przechowawca = znajdz_kolege(wyniki)
nowa_populacja += [populacja[przechowawca]]
# uzupełnienianie populacji randmowymi członkami
while len(nowa_populacja) < rozmiar_populacji:
nowa_populacja += [nowy_osobnik(macierz, start, koniec)]
populacja = copy.deepcopy(nowa_populacja)
ostatnia_odl = odleglosc
najlepsze_wyniki.append([i, odleglosc])
def generate_random_dataset_with_respect_to_categories(df, rows):
df_new = pd.copy(df)
for column in df:
min_cat = min(np.nanmin(df[column]))
max_cat = max(np.nanmax(df[column]))
print('yo')
try:
df_new[column] = np.random.choice(
range(int(min_cat),
int(max_cat) + 1), rows)
except ValueError:
print('hello')
def callback(pd):
global rec_db
global count
pd_temp = []
# print pd
if (count % 4 == 0):
pd_temp = pd.copy(deep=True)
pd_temp = pd_temp.drop('Date', 1)
rec_db = rec_db.append(pd_temp)
count += 1
return pd_temp
def simulate_season(trace):
"""
Simulate a season once, using one random draw from the mcmc chain.
"""
#num_samples = atts.trace().shape[0]
#df_trace = pymc.trace_to_dataframe(trace[:1000])
num_samples = trace['atts'].shape[0]
draw = np.random.randint(0, num_samples)
atts_draw = pd.DataFrame({
'att': trace['atts'][draw, :],
})
defs_draw = pd.DataFrame({
'def': trace['defs'][draw, :],
})
home_draw = trace['home'][draw]
intercept_draw = trace['intercept'][draw]
season = pd.copy()
season = pd.merge(season, atts_draw, left_on='i_home', right_index=True)
season = pd.merge(season, defs_draw, left_on='i_home', right_index=True)
season = season.rename(columns={'att': 'att_home', 'def': 'def_home'})
season = pd.merge(season, atts_draw, left_on='i_away', right_index=True)
season = pd.merge(season, defs_draw, left_on='i_away', right_index=True)
season = season.rename(columns={'att': 'att_away', 'def': 'def_away'})
season['home'] = home_draw
season['intercept'] = intercept_draw
season['home_theta'] = season.apply(lambda x: math.exp(x['intercept'] + x[
'home'] + x['att_home'] + x['def_away']),
axis=1)
season['away_theta'] = season.apply(
lambda x: math.exp(x['intercept'] + x['att_away'] + x['def_home']),
axis=1)
season['home_goals'] = season.apply(
lambda x: np.random.poisson(x['home_theta']), axis=1)
season['away_goals'] = season.apply(
lambda x: np.random.poisson(x['away_theta']), axis=1)
season['home_outcome'] = season.apply(
lambda x: 'win' if x['home_goals'] > x['away_goals'] else 'loss'
if x['home_goals'] < x['away_goals'] else 'draw',
axis=1)
season['away_outcome'] = season.apply(
lambda x: 'win' if x['home_goals'] < x['away_goals'] else 'loss'
if x['home_goals'] > x['away_goals'] else 'draw',
axis=1)
season = season.join(pd.get_dummies(season.home_outcome, prefix='home'))
season = season.join(pd.get_dummies(season.away_outcome, prefix='away'))
return season
def EventHandler(self, pd):
"""Event Handler Object"""
self.pd_temp = pd.copy(deep=True)
#Hardcoded for now
if (pd['Event'].values == 3):
if ((self.buy_db.shape[0] > 0) and (self.buy_db['ItemId'].isin(
self.pd_temp['ItemId']).any() == True)):
index = self.buy_db[self.buy_db['ItemId'].isin(
self.pd_temp['ItemId'])].index
self.buy_db.loc[index, 'Count'] += 1
else:
self.pd_temp['Count'] = 1
self.buy_db = self.buy_db.append(
self.pd_temp.drop(['Date', 'SessionID', 'Event'],
axis=1,
inplace=False))
return
decipline = predicate(
g.subject_objects(
predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/disciple')))
education = predicate(
g.subject_objects(predicate=rdflib.term.URIRef(
u'http://dbpedia.org/ontology/education')))
mentor = predicate(
g.subject_objects(
predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/mentor')))
university = predicate(
g.subject_objects(predicate=rdflib.term.URIRef(
u'http://dbpedia.org/ontology/university')))
#Tensor Factorization by ALS
A, R, _, _, _ = rescal.als([
0.0001 * RDF_Type, 0.0001 * Subject, OWL_SameAs, Influence, BirthPlace,
Occupation, wikiPageExternalLink, Country, profesion, nationality,
deathplace, title, formerteam, league, position, team, termperiod, website,
careerstation, clubs, currentclub, nationalteam, youth, Chancellor, party,
religion, spouse, sucees, battle, militaryrank, relation, branch, rank,
first, last, strokes, collegeteam, associatedband, associatedmusic, genre,
hometown, recordlev, origion, word_net, after, before, predecessor,
successor, language, residence, parent, father, house, issue, mother,
coach, instrument, label, alongside, consitency, deputy, govegene, honpref,
monarch, office, order, prime, movie, education_place, decoration,
decipline, education, mentor, university
], 100)
simstru = cosine_similarity(A, A) #applying cosine similarity
pd = pd.DataFrame(
simstru) #A data structure to identify the location of resources in memory
pd2 = pd.copy()
p_0 = p.copy()
p_0 = p_0.drop(max(p_0.index))
p_0 = p_0.reset_index(drop=True)
p_1 = p.copy()
p_1 = p_1.drop(0).reset_index(drop=True)
x = p_1 / p_0
# plt.plot(x)
# plt.legend(x.keys())
# plt.show()
b = np.full((len(x), d), 1. / d)
s = np.full((len(x), d), 1.)
omega = 30
_x = x.copy()
_x = _x.iloc[omega:]
_x.reset_index(drop=True, inplace=True)
for c in _x.keys():
_x[c] = np.zeros(len(x) - omega)
epsilon = 5
for i in range(omega, len(p)):
print('b=', b[i - omega])
median = geometric_median(p.iloc[0:i],
method='rf_regressor',
options={
'maxiter': 10000,
'tol': 1e-14
})
def combine_columns_ag_news(pd):
pd = pd.copy()
pd[3] = pd[1] + ' ' + pd[2]
del pd[1]
del pd[2]
return pd
def __init__(self, pd, name):
self.pd = pd.copy()
self.name = name
def EventHandler(self, pd):
"""Event Handler Object"""
self.pd_temp = pd.copy(deep=True)
return
residence=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/residence'))) parent=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/parent'))) father=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/father'))) house=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/house'))) issue=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/issue'))) mother=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/mother'))) coach=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/coach'))) instrument=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/instrument'))) label=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/label'))) alongside=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/alongside'))) consitency=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/constituencyMp'))) deputy=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/deputy'))) govegene=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/governorGeneral'))) honpref=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/honorificPrefix'))) monarch=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/monarch'))) office=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/office'))) order=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/order'))) prime=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/property/primeminister'))) movie=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/movie'))) education_place=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/educationPlace'))) decoration=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/decoration'))) decipline=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/disciple'))) education=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/education'))) mentor=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/mentor'))) university=predicate(g.subject_objects(predicate=rdflib.term.URIRef(u'http://dbpedia.org/ontology/university'))) #Tensor Factorization by ALS A, R,_, _, _ = rescal.als([0.0001*RDF_Type,0.0001*Subject,OWL_SameAs,Influence,BirthPlace,Occupation,wikiPageExternalLink,Country,profesion,nationality,deathplace,title,formerteam,league,position,team,termperiod,website,careerstation,clubs,currentclub,nationalteam,youth,Chancellor,party,religion,spouse,sucees,battle,militaryrank,relation,branch,rank,first,last,strokes,collegeteam,associatedband,associatedmusic,genre,hometown,recordlev,origion,word_net,after,before,predecessor,successor,language,residence,parent,father,house,issue,mother,coach,instrument,label,alongside,consitency,deputy,govegene,honpref,monarch,office,order,prime,movie,education_place,decoration,decipline,education,mentor,university],100) simstru=cosine_similarity(A, A)#applying cosine similarity pd=pd.DataFrame(simstru)#A data structure to identify the location of resources in memory pd2=pd.copy()
def limpa_lista(pd):
pd_clear = pd.copy()
for i in range(pd.count()):
pd_clear[i] = 0
return pd_clear