Thanks to @spuri4096 and @chandlergatenbee, the basic cancer stem cell driven model of tumor growth that I implemented in the very first Compute Cancer blog is now programmed in 4 languages:
- C++ (post with the code here)
- MATLAB (post with the code here)
- Java (post with the code here)
- Python (post with the code here)
- Octave – the same code as for MATLAB. Works without any modifications.
- Julia (post with the code here)
Isn’t that the perfect opportunity to see which one is the fastest?
First of all, codes are not completely identical. There are small differences between C++ and MATLAB codes in how the death of a cell is updated, but that shouldn’t affect the comparison too much. The code in Java is much more sophisticated – it uses coded lattice concept and dynamically expanding lattice. Python code (as much as I can see…) does pretty much the same job as the C++ code, so that is the fairest comparison out of all.
I will compare the time needed to simulate cancer from a single cancer stem cell to 100, 250 and 500 thousand of cells. The domain size is set to 1000×1000 grid points (not for Java code, in which lattice expands on demand), so 1 million is the maximal cell number. Of course all simulation parameters are also the same for all codes: probability of symmetric division (ps) = 0.3; proliferation probability (pdiv) = 1/24; migration probability = 15/24; and probability of spontaneous death (alpha) = 0.05. Because of the stochastic nature of the model I will run 20 simulations for each code and report the average time (+- standard deviation).
The results are summarized on the plot below.
As probably everyone expected C++ outcompeted other codes.
Java also did great.
What might surprise some people, MATLAB didn’t perform that bad (it doesn’t suck that bad as people usually think).
The speed of Octave was a surprise for me – I had patience only to run simulations for 100K cell limit…
Of course in each case there is definitely a room for improvement using even more sophisticated language specific programming tricks.
However, from my perspective – if you don’t want to spend too much time on learning programing language, MATLAB (Octave) might be not such a bad idea.
COMPUTE CANCER 
[…] my last post about ABM speed comparison @marcjwills_ suggested that Julia programming language could be a good alternative for slower, […]
LikeLike
Since Java has a native interface (JNI) framework that allows it to call C++ code, it might be interesting to implement a model with JNI and rank it against C++
LikeLike
[…] languages. For example, if you want to oppose my choice of Python in this post then check out his speed comparison of Java, C++, MATLAB, Julia or Python for cellular automatons post for ammunition. Also, let me know what you think of interspersing code within this post. If there […]
LikeLike
Hi there.. I am pretty surprises from these charts… From various benchmarks I did expect Julia to be much closed to C++ (2-4 folds).. if your benchmarks are true there is no reason to use it, speed is only marginally better than Python..
LikeLike
Hi sylvaticus, thanks for the comment! The code written in Julia can be indeed faster – check out the comment made by Marc under the post “Julia – excellent opensource alternative for MATLAB” describing the Julia code that I was using. With the post I wanted to check how big would be the performance increase if not very experienced modeler was to switch to Julia.
LikeLike