ApolloChoiceModelling forum

Hello,

I am estimating a latent class model. Yet, I am running into some problems. My code is working for 2 and 3 classes, although I do get an error message in 3 classes, but my code is not working for 4 and 5 classes.

In the estimation the following error messages are coming up for the 4 and 5 classes.

Testing likelihood function...
Pre-processing likelihood function...
Preparing workers for multithreading...

##################################################################################################################################
This model could potentially be estimated faster using analytical gradients, yet some issue is preventing it from using them. You might want to
ask for help in the Apollo forum (http://www.apollochoicemodelling.com/forum) on how to solve this issue. If you do, please post your code
and data (if not confidential).
##################################################################################################################################

Testing influence of parametersError in checkForRemoteErrors(lapply(cl, recvResult)) :
3 nodes produced errors; first error: invalid (do_set) left-hand side to assignment

The following is what comes up if I use - traceback()

5: stop(count, " nodes produced errors; first error: ", firstmsg,
domain = NA)
4: checkForRemoteErrors(lapply(cl, recvResult))
3: parallel::clusterCall(cl = cl, fun = apollo_parProb, b = b)
2: apollo_logLike(beta1)
1: apollo_estimate(apollo_beta, apollo_fixed, apollo_probabilities,
apollo_inputs)


The message between the ## also show up in the 3 Class model. That is also why I am posting this message to ask for some help and figure out what is going wrong.

If you have any suggestions why this could be and how to solve it, please could you advise.


My code is the following:

### Clear memory
rm(list = ls())

### Load Apollo library
library(apollo)

### Initialise code
apollo_initialise()

### Set core controls
apollo_control = list(
modelName ="LC 4 class",
modelDescr ="LC 4",
indivID ="ID",
nCores = 3, #########Make 4???
noDiagnostics = TRUE

)

# ################################################################# #
#### LOAD DATA AND APPLY ANY TRANSFORMATIONS ####
# ################################################################# #
load("database2.RData")
database = database2

# ################################################################# #
#### DEFINE MODEL PARAMETERS ####
# ################################################################# #

### Vector of parameters, including any that are kept fixed in estimation
apollo_beta = c(
asc_SQ_1 = 0,
asc_SQ_2 = 0,
asc_SQ_3 = 0,
asc_SQ_4 = 0,

beta_intro_lion_a = 0,
beta_intro_lion_b = 0,
beta_intro_lion_c = 0,
beta_intro_lion_d = 0,

beta_intro_che_a = 0,
beta_intro_che_b = 0,
beta_intro_che_c = 0,
beta_intro_che_d = 0,

beta_eco_threat_a = 0,
beta_eco_threat_b = 0,
beta_eco_threat_c = 0,
beta_eco_threat_d = 0,

beta_eco_nonthreat_a = 0,
beta_eco_nonthreat_b = 0,
beta_eco_nonthreat_c = 0,
beta_eco_nonthreat_d = 0,

beta_icon_roan_a = 0,
beta_icon_roan_b = 0,
beta_icon_roan_c = 0,
beta_icon_roan_d = 0,

beta_icon_rhino_a = 0,
beta_icon_rhino_b = 0,
beta_icon_rhino_c = 0,
beta_icon_rhino_d = 0,

beta_icon_tsess_a = 0,
beta_icon_tsess_b = 0,
beta_icon_tsess_c = 0,
beta_icon_tsess_d = 0,

beta_price_a = 0,
beta_price_b = 0,
beta_price_c = 0,
beta_price_d = 0,

delta_a = 0,
delta_b = 0,
delta_c = 0,
delta_d = 0
)
### Vector with names (in quotes) of parameters to be kept fixed at their starting value in apollo_beta, use apollo_beta_fixed = c() if none
apollo_beta_fixed = c()

apollo_fixed = c("delta_a")

# ################################################################# #
#### DEFINE LATENT CLASS COMPONENTS ####
# ################################################################# #

apollo_lcPars=function(apollo_beta, apollo_inputs){
lcpars = list()

lcpars[["asc_SQ"]] = list(asc_SQ_1, asc_SQ_2, asc_SQ_3, asc_SQ_4)
lcpars[["beta_intro_lion"]] = list(beta_intro_lion_a, beta_intro_lion_b, beta_intro_lion_c, beta_intro_lion_d)
lcpars[["beta_intro_che"]] = list(beta_intro_che_a, beta_intro_che_b, beta_intro_che_c, beta_intro_che_d)
lcpars[["beta_eco_threat"]] = list(beta_eco_threat_a, beta_eco_threat_b, beta_eco_threat_c, beta_eco_threat_d)
lcpars[["beta_eco_nonthreat"]] = list(beta_eco_nonthreat_a, beta_eco_nonthreat_b, beta_eco_nonthreat_c, beta_eco_nonthreat_d)
lcpars[["beta_icon_roan"]] = list(beta_icon_roan_a, beta_icon_roan_b, beta_icon_roan_c, beta_icon_roan_d)
lcpars[["beta_icon_rhino"]] = list(beta_icon_rhino_a, beta_icon_rhino_b, beta_icon_rhino_c, beta_icon_rhino_d)
lcpars[["beta_icon_tsess"]] = list(beta_icon_tsess_a, beta_icon_tsess_b, beta_icon_tsess_c, beta_icon_tsess_d)
lcpars[["beta_price"]] = list(beta_price_a, beta_price_b, beta_price_c, beta_price_d)


V=list()
V[["class_a"]] = delta_a
V[["class_b"]] = delta_b
V[["class_c"]] = delta_c
V[["class_d"]] = delta_d

mnl_settings = list(
alternatives = c(class_a=1, class_b=2,class_c=3, class_d=4),
avail = 1,
choiceVar = NA,
V = V
)
lcpars[["pi_values"]] = apollo_mnl(mnl_settings, functionality="raw")

lcpars[["pi_values"]] = apollo_firstRow(lcpars[["pi_values"]], apollo_inputs)

return(lcpars)
}

# ################################################################# #
#### GROUP AND VALIDATE INPUTS ####
# ################################################################# #

apollo_inputs = apollo_validateInputs()

# ################################################################# #
#### DEFINE MODEL AND LIKELIHOOD FUNCTION ####
# ################################################################# #

apollo_probabilities=function(apollo_beta, apollo_inputs, functionality="estimate"){

### Attach inputs and detach after function exit
apollo_attach(apollo_beta, apollo_inputs)
on.exit(apollo_detach(apollo_beta, apollo_inputs))

### Create list of probabilities P
P = list()

### Define settings for MNL model component that are generic across classes
mnl_settings = list(
alternatives = c(one=1, two=2, SQ=3),
avail = list(one=1, two=1, SQ=1),
choiceVar = Choice
)

### Loop over classes
for(s in 1:4){


### Compute class-specific utilities
V=list()

V[['one']] =
beta_intro_lion[[s]]*alt1_car_lion +
beta_intro_che[[s]]*alt1_car_che +
beta_eco_threat[[s]]*alt1_eco_threat +
beta_eco_nonthreat[[s]]*alt1_eco_nonthreat +
beta_icon_roan[[s]]*alt1_endg_roan +
beta_icon_rhino[[s]]*alt1_endg_rhino +
beta_icon_tsess[[s]]*alt1_endg_tsess +
beta_price[[s]]*alt1.price

V[['two']] =
beta_intro_lion[[s]]*alt2_car_lion +
beta_intro_che[[s]]*alt2_car_che +
beta_eco_threat[[s]]*alt2_eco_threat +
beta_eco_nonthreat[[s]]*alt2_eco_nonthreat +
beta_icon_roan[[s]]*alt2_endg_roan +
beta_icon_rhino[[s]]*alt2_endg_rhino +
beta_icon_tsess[[s]]*alt2_endg_tsess +
beta_price[[s]]*alt2.price

V[['SQ']] = asc_SQ[[s]] + beta_price[[s]]* sq.price


mnl_settings$V = V
mnl_settings$componentName = paste0("Class_",s)

### Compute within-class choice probabilities using MNL model
P[[paste0("Class_",s)]] = apollo_mnl(mnl_settings, functionality)

### Take product across observation for same individual
P[[paste0("Class_",s)]] = apollo_panelProd(P[[paste0("Class_",s)]], apollo_inputs ,functionality)

s=s+1
}

### Compute latent class model probabilities
lc_settings = list(inClassProb = P, classProb=pi_values)
P[["model"]] = apollo_lc(lc_settings, apollo_inputs, functionality)

### Prepare and return outputs of function
P = apollo_prepareProb(P, apollo_inputs, functionality)
return(P)
}

# ################################################################# #
#### MODEL ESTIMATION ####
# ################################################################# #

### Optional starting values search
#apollo_beta=apollo_searchStart(apollo_beta, apollo_fixed,apollo_probabilities, apollo_inputs)

model = apollo_estimate(apollo_beta, apollo_fixed, apollo_probabilities, apollo_inputs)

traceback()
#Output to screen
apollo_modelOutput(model)



And my Output is the above as mentioned.


Thank you very much

Hanno

Hanno

could you please first confirm what version of Apollo you are using?

You are also still using the old Apollo syntax for latent class, please see the new examples on the website

Thanks

Stephane