Development of Regression Models considering Time-Lag and Aerosols for Predicting Heating Loads in Buildings

Table 21

Residual statistics for a DOE large office building depending on time-lag.

Residual statistics^a
		Minimum	Maximum	Mean	Std. deviation	N
Time-lag0	Predicted value	−767601856.00	432252768.00	63956011.14	103525710.371	20,352
	Residual	−392874016.000	1979168896.000	0.000	172856439.204	20,352
	Std. predicted value	−8.032	3.558	0.000	1.000	20,352
	Std. residual	−2.272	11.447	0.000	1.000	20,352
Time-lag1	Predicted value	−684187840.00	516718464.00	63956412.79	103925691.005	20,351
	Residual	−356752608.000	2007745792.000	0.000	172622025.484	20,351
	Std. predicted value	−7.199	4.357	0.000	1.000	20,351
	Std. residual	−2.066	11.628	0.000	1.000	20,351
Time-lag2	Predicted value	−47442988.00	48511468.00	5299326.20	8882662.010	20,350
	Residual	−35604016.000	136248016.000	0.000	12514869.803	20,350
	Std. predicted value	−5.938	4.865	0.000	1.000	20,350
	Std. residual	−2.844	10.884	0.000	1.000	20,350

^aDependent variable: heating load of a DOE large office building.