def baseline_model_billing_single_record(il_electricity_cdd_hdd_billing_monthly):
baseline_meter_data = il_electricity_cdd_hdd_billing_monthly["meter_data"][-2:]
temperature_data = il_electricity_cdd_hdd_billing_monthly["temperature_data"]
blackout_start_date = il_electricity_cdd_hdd_billing_monthly["blackout_start_date"]
baseline_data = create_caltrack_billing_design_matrix(
baseline_meter_data, temperature_data
)
model_results = fit_caltrack_usage_per_day_model(
baseline_data, use_billing_presets=True, weights_col="n_days_kept"
)
return model_results
def reporting_model_billing(il_electricity_cdd_hdd_billing_monthly):
meter_data = il_electricity_cdd_hdd_billing_monthly["meter_data"]
meter_data.value = meter_data.value - 50
temperature_data = il_electricity_cdd_hdd_billing_monthly[
"temperature_data"]
blackout_start_date = il_electricity_cdd_hdd_billing_monthly[
"blackout_start_date"]
baseline_meter_data, warnings = get_baseline_data(meter_data,
end=blackout_start_date)
baseline_data = create_caltrack_billing_design_matrix(
baseline_meter_data, temperature_data)
model_results = fit_caltrack_usage_per_day_model(baseline_data,
use_billing_presets=True,
weights_col="n_days_kept")
return model_results
def reporting_model_daily(reporting_data_daily):
model_results = fit_caltrack_usage_per_day_model(reporting_data_daily)
return model_results
def baseline_model_daily(baseline_data_daily):
model_results = fit_caltrack_usage_per_day_model(baseline_data_daily)
return model_results
def test_total_average_metrics():
data = pd.DataFrame(
{
"meter_value": [6, 1, 1, 6],
"cdd_65": [5, 0, 0.1, 0],
"hdd_65": [0, 0.1, 0.1, 5],
"start": pd.date_range(start="2016-01-02", periods=4, freq="D", tz="UTC"),
}
).set_index("start")
model_results = fit_caltrack_usage_per_day_model(data, fit_intercept_only=True)
json_result = model_results.json()
totals_metrics = json_result["totals_metrics"]
assert round(totals_metrics["observed_length"], 3) == 3.000
assert round(totals_metrics["predicted_length"], 3) == 3.000
assert round(totals_metrics["merged_length"], 3) == 3.000
assert round(totals_metrics["num_parameters"], 3) == 0
assert round(totals_metrics["observed_mean"], 3) == 2.667
assert round(totals_metrics["predicted_mean"], 3) == 3.5
assert round(totals_metrics["observed_variance"], 3) == 5.556
assert round(totals_metrics["predicted_variance"], 3) == 0
assert round(totals_metrics["observed_skew"], 3) == 1.732
assert round(totals_metrics["predicted_skew"], 3) == 0
assert round(totals_metrics["observed_cvstd"], 3) == 1.083
assert round(totals_metrics["predicted_cvstd"], 3) == 0
assert round(totals_metrics["rmse"], 3) == 2.5
assert round(totals_metrics["rmse_adj"], 3) == 2.5
assert round(totals_metrics["cvrmse"], 3) == 0.938
assert round(totals_metrics["cvrmse_adj"], 3) == 0.938
assert round(totals_metrics["mape"], 3) == 1.806
assert round(totals_metrics["mape_no_zeros"], 3) == 1.806
assert round(totals_metrics["num_meter_zeros"], 3) == 0
assert round(totals_metrics["nmae"], 3) == 0.938
assert round(totals_metrics["nmbe"], 3) == 0.312
assert round(totals_metrics["confidence_level"], 3) == 0.9
assert round(totals_metrics["single_tailed_confidence_level"], 3) == 0.95
assert totals_metrics["observed_kurtosis"] is None
assert totals_metrics["predicted_kurtosis"] is None
assert totals_metrics["r_squared"] is None
assert totals_metrics["r_squared_adj"] is None
assert totals_metrics["autocorr_resid"] is None
assert totals_metrics["n_prime"] is None
assert totals_metrics["degrees_of_freedom"] is None
assert totals_metrics["t_stat"] is None
assert totals_metrics["cvrmse_auto_corr_correction"] is None
assert totals_metrics["approx_factor_auto_corr_correction"] is None
assert totals_metrics["fsu_base_term"] is None
json_result = model_results.json()
avgs_metrics = json_result["avgs_metrics"]
assert round(avgs_metrics["observed_length"], 3) == 4.000
assert round(avgs_metrics["predicted_length"], 3) == 4.000
assert round(avgs_metrics["merged_length"], 3) == 4.000
assert round(avgs_metrics["num_parameters"], 3) == 0
assert round(avgs_metrics["observed_mean"], 3) == 3.5
assert round(avgs_metrics["predicted_mean"], 3) == 3.5
assert round(avgs_metrics["observed_variance"], 3) == 6.25
assert round(avgs_metrics["predicted_variance"], 3) == 0
assert round(avgs_metrics["observed_skew"], 3) == 0
assert round(avgs_metrics["predicted_skew"], 3) == 0
assert round(avgs_metrics["observed_cvstd"], 3) == 0.825
assert round(avgs_metrics["predicted_cvstd"], 3) == 0
assert round(avgs_metrics["observed_kurtosis"], 3) == -6.0
assert round(avgs_metrics["predicted_kurtosis"], 3) == 0
assert round(avgs_metrics["rmse"], 3) == 2.5
assert round(avgs_metrics["rmse_adj"], 3) == 2.5
assert round(avgs_metrics["cvrmse"], 3) == 0.714
assert round(avgs_metrics["cvrmse_adj"], 3) == 0.714
assert round(avgs_metrics["mape"], 3) == 1.458
assert round(avgs_metrics["mape_no_zeros"], 3) == 1.458
assert round(avgs_metrics["num_meter_zeros"], 3) == 0
assert round(avgs_metrics["nmae"], 3) == 0.714
assert round(avgs_metrics["nmbe"], 3) == 0
assert round(avgs_metrics["confidence_level"], 3) == 0.9
assert round(avgs_metrics["n_prime"], 3) == 12.0
assert round(avgs_metrics["single_tailed_confidence_level"], 3) == 0.95
assert round(avgs_metrics["autocorr_resid"], 3) == -0.5
assert round(avgs_metrics["degrees_of_freedom"], 3) == 12.0
assert round(avgs_metrics["t_stat"], 3) == 1.782
assert round(avgs_metrics["cvrmse_auto_corr_correction"], 3) == 0.577
assert round(avgs_metrics["approx_factor_auto_corr_correction"], 3) == 1.08
assert round(avgs_metrics["fsu_base_term"], 3) == 0.794
assert avgs_metrics["r_squared"] is None
assert avgs_metrics["r_squared_adj"] is None
