Tom Slater's peer review: output analysis

[amyheather]

Peer review from @tbslater

Initialisation Bias

• Typo…

• This needs rewriting…

• Might be worth explaining what these functions are doing…

• I think the following function would benefit from some line by line commenting as it is not always clear how each function is manipulating the data. --> Function has quite alot of comments and then broken down line by line with markdown explanations, so will not add any further for this.

Just scrolled down and the explanation of warm-up function helps. But I’d still add some more commenting 😊

Performance measures

I like the comparisons drawn between DES and queueing theory (without going into any unnecessary detail).
The utilisation function is very well explained.

• Note: the average would be the total area / time elapsed. But I like the visual demonstration – it helps cement the idea.

• A summary of all the performance metrics at the end might be useful.

Replications

• Going back to randomness and the associated plot, I think this plot and explanation better represents the impact of randomness. Something similar earlier might be beneficial

• Maybe include some performance metrics in the code output (e.g. utilisation).

Length of Warm-Up

Nice plot showing steady-state behaviour.

• How much do the results change if you move the warm-up cutoff forwards or backwards? Could you show these results?

Number of Replications

• Manual inspection – I think this needs a little more explaining (i.e. what all the output is showing).

• Automated – so do you just select the highest number of runs (i.e. 19 in this case)? I’d like a sentence which uses the output to state the amount of replications you would use for this model.

Parallel Processing

Nice and clear 😊