Alberto
2018-02-18 10:22:26 UTC
Hola a todos,
cada vez que intento ejecutar el código adjuntado cuando llego al comando de predict() salta un pop-up de R diciendo 'R encountered a fatal error'. No se donde puede estar el fallo y no entiendo tampoco por qué pasa esto puesto que a mi parecer el código está bien. Alguna idea de cual puede ser el problema?
Muchas gracias!
set.seed(6)
model <- init.liquidSVM(Casualty_Severity~Vehicle_Type+Vehicle_Manoeuvre+Junction_Location+Skidding_and_Overturning+Hit_Object_off_Carriageway+First_Point_of_Impact+Journey_Purpose_of_Driver+Sex_of_Driver+Age_Band_of_Driver+Propulsion_Code+Age_of_Vehicle+Driver_Home_Area_Type+Sex_of_Casualty+Age_Band_of_Casualty+Car_Passenger+Casualty_Type+Number_of_Vehicles+Hour_of_Day+First_Road_Class+Road_Type+Speed_limit+Junction_Detail+Light_Conditions+Weather_Conditions+Road_Surface_Conditions+Urban_or_Rural_Area+month+other_vehic, trainning, threads = -1, gammas = c(0.0001,0.001,0.04,0.1,50, 1,10, 100), c_values = c(0.00001, 0.0001, 0.001,1,10,25,50))
trainSVMs(model, threads = -1, gammas = c(0.0001,0.001,0.04,0.1,50, 1,10, 100), c_values = c(0.00001, 0.0001, 0.001,1,10,25,50), solver = 'ls', command.args=list(L=2, T=-1, d=1))
selectSVMs(model)
svm.probs <- predict(model, type = 'response', newdata = tst)
[[alternative HTML version deleted]]
cada vez que intento ejecutar el código adjuntado cuando llego al comando de predict() salta un pop-up de R diciendo 'R encountered a fatal error'. No se donde puede estar el fallo y no entiendo tampoco por qué pasa esto puesto que a mi parecer el código está bien. Alguna idea de cual puede ser el problema?
Muchas gracias!
set.seed(6)
model <- init.liquidSVM(Casualty_Severity~Vehicle_Type+Vehicle_Manoeuvre+Junction_Location+Skidding_and_Overturning+Hit_Object_off_Carriageway+First_Point_of_Impact+Journey_Purpose_of_Driver+Sex_of_Driver+Age_Band_of_Driver+Propulsion_Code+Age_of_Vehicle+Driver_Home_Area_Type+Sex_of_Casualty+Age_Band_of_Casualty+Car_Passenger+Casualty_Type+Number_of_Vehicles+Hour_of_Day+First_Road_Class+Road_Type+Speed_limit+Junction_Detail+Light_Conditions+Weather_Conditions+Road_Surface_Conditions+Urban_or_Rural_Area+month+other_vehic, trainning, threads = -1, gammas = c(0.0001,0.001,0.04,0.1,50, 1,10, 100), c_values = c(0.00001, 0.0001, 0.001,1,10,25,50))
trainSVMs(model, threads = -1, gammas = c(0.0001,0.001,0.04,0.1,50, 1,10, 100), c_values = c(0.00001, 0.0001, 0.001,1,10,25,50), solver = 'ls', command.args=list(L=2, T=-1, d=1))
selectSVMs(model)
svm.probs <- predict(model, type = 'response', newdata = tst)
[[alternative HTML version deleted]]