def test_time_series_instance_jobs():
""" Test that you can instantiate a series. """
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
assert isinstance(ts, stax.TimeSeries)
def test_bad_time_series_instance():
""" Test that index of series has to be a datetime """
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = range(df.shape[0])
JOBS = df.Jobs
try:
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
except Exception as e:
assert isinstance(e, AssertionError)
def test_ACF_jobs():
"""Test ACF and PACF functionality."""
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
ACF = stax.tools.ACF(ts)
PACF = stax.tools.PACF(ts)
assert len(PACF) == 41
assert len(ACF) == 41
def test_seasonal_decomp_jobs():
"""Test a seasonal decomposition."""
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
decomp = stax.tools.decompose_series(ts)
assert decomp["method"] == "multiplicative"
assert isinstance(decomp["trend"], list)
assert isinstance(decomp["seasonal"], list)
assert isinstance(decomp["resid"], list)
def test_TBATS_model_jobs():
"""Test TBATS model training."""
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
model, test_pred, test_conf, test_metrics, OOS_pred, OOS_conf = stax.train_tbats(
ts)
assert model.aic < 1900
assert isinstance(test_conf, list)
assert test_metrics[0]['mean_absolute_percent_error'] < 0.05
assert len(OOS_pred) == 12
assert len(test_pred) == 18
def test_ARIMA_model_jobs():
"""Test ARIMA model training. l"""
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
model, test_pred, test_conf, test_metrics, OOS_pred, OOS_conf = stax.train_arima(
ts)
assert model.aic() < 1800
assert len(test_conf) == 18
assert test_metrics[0]['mean_absolute_percent_error'] < 0.05
assert len(OOS_pred) == 12
assert len(test_pred) == 18
def test_train_all_models_jobs():
"""Test .train_models() method and .calculate_statistics()."""
df = pd.read_csv('../data/jobs-trend.csv').set_index("Date")
df.index = pd.to_datetime(df.index)
JOBS = df.Jobs
ts = stax.TimeSeries(JOBS, "monthly", 0.8)
ts.train_models()
ts.calculate_statistics()
assert sorted(list(ts.experiment_results["models"].keys())) == [
'ARIMA', 'ExponentialSmoothing', 'TBATS'
]
assert sorted(ts.experiment_results["seasonal_decomposition"].keys()) == [
'method', 'resid', 'seasonal', 'trend'
]
assert sorted(
ts.experiment_results["autocorrelation"].keys()) == ['ACF', 'PACF']