def print_point_statistics(data,
models,
externalmodels=None,
externalforecasts=None,
indexers=None):
ret = "Model & Order & RMSE & SMAPE & Theil's U \\\\ \n"
for count, model in enumerate(models, start=0):
_rmse, _smape, _u = Measures.get_point_statistics(
data, model, indexers)
ret += model.shortname + " & "
ret += str(model.order) + " & "
ret += str(_rmse) + " & "
ret += str(_smape) + " & "
ret += str(_u)
#ret += str(round(Measures.TheilsInequality(np.array(data[fts.order:]), np.array(forecasts[:-1])), 4))
ret += " \\\\ \n"
if externalmodels is not None:
l = len(externalmodels)
for k in np.arange(0, l):
ret += externalmodels[k] + " & "
ret += " 1 & "
ret += str(round(Measures.rmse(data, externalforecasts[k][:-1]),
2)) + " & "
ret += str(
round(Measures.smape(data, externalforecasts[k][:-1]),
2)) + " & "
ret += str(
round(Measures.UStatistic(data, externalforecasts[k][:-1]), 2))
ret += " \\\\ \n"
print(ret)
def print_point_statistics(data, models, externalmodels = None, externalforecasts = None, indexers=None):
"""
Run point benchmarks on given models and data and print the results
:param data: test data
:param models: a list of FTS models to benchmark
:param externalmodels: a list with benchmark models (façades for other methods)
:param externalforecasts:
:param indexers:
:return:
"""
ret = "Model & Order & RMSE & SMAPE & Theil's U \\\\ \n"
for count,model in enumerate(models,start=0):
_rmse, _smape, _u = Measures.get_point_statistics(data, model, indexers)
ret += model.shortname + " & "
ret += str(model.order) + " & "
ret += str(_rmse) + " & "
ret += str(_smape)+ " & "
ret += str(_u)
#ret += str(round(Measures.TheilsInequality(np.array(data[fts.order:]), np.array(forecasts[:-1])), 4))
ret += " \\\\ \n"
if externalmodels is not None:
l = len(externalmodels)
for k in np.arange(0,l):
ret += externalmodels[k] + " & "
ret += " 1 & "
ret += str(round(Measures.rmse(data, externalforecasts[k][:-1]), 2)) + " & "
ret += str(round(Measures.smape(data, externalforecasts[k][:-1]), 2))+ " & "
ret += str(round(Measures.UStatistic(data, externalforecasts[k][:-1]), 2))
ret += " \\\\ \n"
print(ret)
def forecast_params(data, train_split, method, params, plot=False):
train, test = sampling.train_test_split(data, train_split)
fcst = method(train, test, params)
_output = params['output']
_step = params.get('step', 1)
_offset = params['order'] + _step - 1
yobs = test[_output].iloc[_offset:].values
if plot:
plt.figure(figsize=(20, 10))
plt.plot(yobs)
plt.plot(fcst)
plt.show()
rmse = Measures.rmse(yobs, fcst)
print("RMSE: ", rmse)
nrmse = metrics.normalized_rmse(yobs, fcst)
print("nRMSE: ", nrmse)
smape = Measures.smape(yobs, fcst)
print("SMAPE: ", smape)
u = Measures.UStatistic(yobs, fcst)
print("U Statistic: ", u)
return rmse, nrmse, smape, u
def forecast_best_params(data,
train_split,
method_id,
method,
space,
plot=False,
save=False):
print("Running experiment ", method_id)
best = pickle.load(open("best_" + method_id + ".pkl", "rb"))
train, test = sampling.train_test_split(data, train_split)
best_params = space_eval(space, best)
fcst = method(train, test, best_params)
_order = best_params['order']
_output = best_params['output']
yobs = test[_output].iloc[_order:].values
if plot:
plt.figure(figsize=(20, 10))
plt.plot(yobs)
plt.plot(fcst)
plt.show()
rmse = Measures.rmse(yobs, fcst)
print("RMSE: ", rmse)
nrmse = metrics.normalized_rmse(yobs, fcst)
print("nRMSE: ", nrmse)
smape = Measures.smape(yobs, fcst)
print("SMAPE: ", smape)
u = Measures.UStatistic(yobs, fcst)
print("U Statistic: ", u)
if save:
results = {
"method_id": method_id,
"forecast": fcst,
"RMSE": rmse,
"SMAPE": smape,
"U": u
}
pickle.dump(results, open("results_" + method_id + ".pkl", "wb"))
return rmse, nrmse, smape, u
def rolling_window_forecast_params(data, train_percent, window_size, method,
params):
# get training days
training_days = pd.unique(data.index.date)
fcst = []
yobs = []
for day in training_days:
print("Processing :", day)
daily_data = data[data.index.date == day]
nsamples = len(daily_data.index)
train_size = round(nsamples * train_percent)
test_end = 0
index = 0
while test_end < nsamples:
train_start, train_end, test_start, test_end = get_data_index(
index, train_size, window_size, nsamples)
train = data[train_start:train_end]
test = data[test_start:test_end]
index += window_size
f = method(train, test, params)
fcst.extend(f)
_step = params.get('step', 1)
_output = params['output']
_offset = params['order'] + _step - 1
yobs.extend(test[_output].iloc[_offset:].values)
rmse = Measures.rmse(yobs, fcst)
print("RMSE: ", rmse)
nrmse = metrics.normalized_rmse(yobs, fcst)
print("nRMSE: ", nrmse)
smape = Measures.smape(yobs, fcst)
print("SMAPE: ", smape)
u = Measures.UStatistic(yobs, fcst)
print("U Statistic: ", u)
return rmse, nrmse, smape, u