Dear stan users:

I am using Rstan to run a relatively complex model in a cluster. I initially set the wall times to be 8 hours but after 8 hours of waiting, my job got killed and I looked into the log file the job was only 40% complete and exceeded the set wall time.

Gradient evaluation took 0.05 seconds

1000 transitions using 10 leapfrog steps per transition would take 500 seconds.

Adjust your expectations accordingly!

**Iteration: 1 / 500 [ 0%] (Warmup)**

**Iteration: 1 / 500 [ 0%] (Warmup)**

**Iteration: 1 / 500 [ 0%] (Warmup)**

**Iteration: 50 / 500 [ 10%] (Warmup)**

**Iteration: 50 / 500 [ 10%] (Warmup)**

**Iteration: 50 / 500 [ 10%] (Warmup)**

**Iteration: 50 / 500 [ 10%] (Warmup)**

**Iteration: 100 / 500 [ 20%] (Warmup)**

**Iteration: 100 / 500 [ 20%] (Warmup)**

**Iteration: 100 / 500 [ 20%] (Warmup)**

**Iteration: 100 / 500 [ 20%] (Warmup)**

**Iteration: 150 / 500 [ 30%] (Warmup)**

**Iteration: 150 / 500 [ 30%] (Warmup)**

**Iteration: 150 / 500 [ 30%] (Warmup)**

**Iteration: 150 / 500 [ 30%] (Warmup)**

**Iteration: 200 / 500 [ 40%] (Warmup)**

**Iteration: 200 / 500 [ 40%] (Warmup)**

**Iteration: 200 / 500 [ 40%] (Warmup)**

**Iteration: 200 / 500 [ 40%] (Warmup)**

**=>> PBS: job killed: walltime 28836 exceeded limit 28800**

My model runs 2000 iterations with 500 warmup per chain for 4 chains and I would like to know are there any ways that I can

a) save the first 40% of the simulation outputs?

b) perhaps speed up the programme?

I am very new to HPC and stuff, any ideas would be greatly appreciated.

I have attached my Rscript, stan code and pbs file in this post.

Thanks in advance,

Jaslene2_comp_full_classical.r (756 Bytes)

full_state_space_true_classical.stan (5.5 KB)

2_comp_full_classical.txt (274 Bytes)