How to combine the models if any one of them is showing NULL value in FABLE package?

[Adarsha1991]

Code

fit1 <- train %>%
model(
ets = ETS(value),
arima = ARIMA(value),
SNAIVE= SNAIVE(value ~ lag("year")),
prophet = prophet(value),
nn = NNETAR(value),
var=VAR(value ~ AR(4)),
tslm=TSLM(value ~ trend() + season())

) %>%
mutate(
mixed = (ets + arima + tslm + prophet + nn + var)/6
)

fc1 <- fit1 %>% forecast(h = "1 years")

#for example if ARIMA model is NULL then mixed values in fc1 would not throw any result. We need mixed model value in fc1 that excludes NULL model and give us results. Can anyone help us? we are using Fable package.