# selektuj uset/activity
def find_optimal_HMM(activities, cluster_range):
scores=[]
for k in cluster_range:
model=GaussianHMM(n_components=k, covariance_type="diag", n_iter=1000).fit(activities)
log_likelihood=model.score(activities)
bic=2*log_likelihood-k*np.log(len(activities))
scores.append((k, log_likelihood, bic))
return scores
model.score()
### Print or save clusters
#for result in results_single_hmm:
# draw_clusters(result)
### Ploting complete behavior
'''
# %%
df['mempool-size'].iloc[1000:].plot()
# %% sklearn linear regression
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
nlag = 10
cutoff = 1400
model = LinearRegression()
feature_names = [
'mempool-size', 'n-transactions-per-block', 'n-transactions', 'hash-rate'
]
target_name = ['transaction-fees']
features = []
for lag in range(nlag):
features.append(df[feature_names].shift(lag))
features[-1].columns = [name + str(lag) for name in feature_names]
features = pd.concat(features, axis=1)
#features.fillna(method='backfill', axis=0, inplace=True)
features = features.loc[cutoff:, :]
target = df.loc[1200:, target_name]
#target = target.loc[cutoff:, :]
X_train, X_test, y_train, y_test = train_test_split(features,
target,
test_size=0.1)
model.fit(X_train, y_train)
print(model.score(X_test, y_test))
# In[74]:
plt.figure(figsize=(15, 5))
plt.title("Linear Regression Model")
plt.xlabel("Date")
plt.ylabel("Open Price")
plt.plot(train.index, y_train, "blue", label='Training data')
plt.plot(test.index, y_test, "green", label='Testing data')
plt.plot(test.index, y_pred, "red", label='Predicted data')
plt.xticks(np.arange(0, 242, 25), dataset['Date'][0:242:25])
#plt.plot(test.index,regressor.predict(x_test),"yellow")
plt.legend()
# In[83]:
r_sq = model.score(x_train, y_train)
print('coefficient of determination:', r_sq)
print('Slope of model: ', model.coef_)
print('Intercept of model: ', model.intercept_)
# In[57]:
err = metrics.mean_squared_error(y_test, y_pred)
print("Mean squared error: ", err)
# In[85]:
# k nearest neighbour
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
X = series
size = int(len(X) * 0.66)
train, test = X[0:size], X[size:len(X)]
dates_test=dates[size:len(dates)]
history = [x for x in train]
predictions = list()
for t in range(len(test)):
model = ARIMA(history, order=(3,1,0))
model_fit = model.fit(disp=0)
output = model_fit.forecast()
yhat = output[0]
predictions.append(yhat)
obs = test[t]
history.append(obs)
print('predicted=%f, expected=%f, average= %f ' % (yhat, obs, obs-yhat))
score=model.score(test)
print(score)
# plot
pyplot.plot(dates_test,test, lw=1.5, color="blue", label="excepted")
pyplot.plot(dates_test,predictions, color='red', label="predicted")
pyplot.show()
mse = mean_squared_error(test, predictions)
print("Mean Squared Error:",mse)
rmse = math.sqrt(mse)
print("Root Mean Squared Error:", rmse)
predictions = list()
yhat = float(model_fit.forecast()[0])
yhat = bias + inverse_difference(history, yhat, months_in_year)
predictions.append(yhat)
history.append(y[0])
print('>Predicted=%.3f, Expected=%3.f' % (yhat, y[0]))
# rolling forecasts
# predict
# difference data
months_in_year = 12
diff = difference(history, months_in_year)
model = ARIMA(diff, order=(0, 0, 1))
model_fit = model.fit(trend='nc', disp=0)
for i in range(1, len(y)):
yhat = model_fit.forecast()[0]
yhat = bias + inverse_difference(history, yhat, months_in_year)
predictions.append(yhat)
# observation
obs = y[i]
history.append(obs)
print('>Predicted=%.3f, Expected=%3.f' % (yhat, obs))
# report performance
mse = mean_squared_error(y, predictions)
rmse = sqrt(mse)
print('RMSE: %.3f' % rmse)
print(model.score(y, predictions))
pyplot.plot(y, color='green')
pyplot.plot(predictions, color='red')
pyplot.show()
