Mass Rapid Transit Ridership Forecast Based on Direct Ridership Models: A Case Study in Wuhan, China

Table 5

Eigenvectors of the principal components and the correlation matrix.

Principal components	Variables	Component 1	Component 2	Component 3	Component 4
Component 1: factors related to the built environment	Restaurant_Num	0.837
	Shopping_Num	0.791
	Financial_Num	0.774
	Com_Area	0.768
	Hotel_Num	0.716
	Parking_Num	0.716
	Recreational_Num	0.715
	Offi_Area	0.679
	Land_use_mix	0.629
Component 2: factors related to jobs-housing spatial structure	P_Accessibility		0.889
	J_Accessibility		0.859
	Dis_to_centers		−0.781
	Bus_line_Num		0.681
	Population		0.613
	Dummy_CBD		0.564
	Employment		0.510
Component 3: factors related to station attributes	Dummy_line_transfer			0.833
Component 4: factors related to large compound	College_Num				0.734
	Hospital_Num				0.595

Correlation coefficients between variables and components greater than 0.5 are shown in the table.