I used a lot of different tools and strategies to make this paper easily reproducible at different levels. There's Docker container for the highest level of reproducibility, and package versions are managed with renv. The data used in the paper is hosted on Zenodo to avoid long queue times when downloading from the Climate Data Store and future-proof for when it goes away and checksumed before using it.
This paper presents a fine example of high-throughput computational materials screening studies, mainly focusing on the carbon nanoclusters of different sizes. In the paper, a set of diverse empirical and machine-learned interatomic potentials, which are commonly used to simulate carbonaceous materials, is benchmarked against the higher-level density functional theory (DFT) data, using a range of diverse structural features as the comparison criteria. Trying to reproduce the data presented here (even if you only consider a subset of the interaction potentials) will help you devise an understanding as to how you could approach a high-throughput structure prediction problem. Even though we concentrate here on isolated/finite nanoclusters, AIRSS (and other similar approaches like USPEX, CALYPSO, GMIN, etc.,) can also be used to predict crystal structures of different class of materials with applications in energy storage, catalysis, hydrogen storage, and so on.
The negative surface enthalpies in figure 5 are surprising. At least one group has attempted to reproduce these using a different code and obtained positive enthalpies. This was attributed to the inability of that code to independently relax the three simulation cell vectors resulting in an unphysical water density. This demonstrates how sensitive these results can be to the particular implementation of simulation algorithms in different codes. Similarly the force field used is now very popular. Its functional form and full set of parameters can be found in the literature. However differences in how different simulation codes implement truncation, electrostatics etc can lead to significant difference in results such as these. It would be a valuable exercise to establish if exactly the same force field as that used here can be reproduced from only its specification in the literature. The interfacial energies of interest should be reproducible with simulations on modest numbers of processors (a few dozen) with run times for each being 1-2 days. Each surface is an independent calculation and so these can be run concurrently during the ReproHack.
This paper proposes a probabilistic planner that can solve goal-conditional tasks such as complex continuous control problems. The approach reaches state-of-the-art performance when compared to current deep reinforcement learning algorithms. However, the method relies on an ensemble of deep generative models and is computationally intensive. It would be interesting to reproduce the results presented in this paper on their robotic manipulation and navigation problems as these are very challenging problems that current reinforcement learning methods cannot easily solve (and when they do, they require a significantly larger number of experiences). Can the results be reproduced out-of-the-box with the provided code?
In theory, reproducing this paper should only require a clone of a public Git repository, and the execution of a Makefile (detailed in the README of the paper repository at https://github.com/psychoinformatics-de/paper-remodnav). We've set up our paper to be dynamically generated, retrieving and installing the relevant data and software automatically, and we've even created a tutorial about it, so that others can reuse the same setup for their work. Nevertheless, we've for example never tried it out across different operating systems - who knows whether it works on Windows? We'd love to share the tips and tricks we found to work, and even more love feedback on how to improve this further.
Most of the material is available through Jupyter notebooks in GitHub, and it should be easy to reproduce with the help of Binder. With the notebooks, you could experiment with different parameters to the ones analyzed in the paper. It also contains a large dataset of physical parameters of galaxies analysed in this work. We expect this work to be easily reproducible in the steps described in the repository.
I suggested a few papers last year. I’m hoping that we’ve improved our reproducibility with this one, this year. We’ve done our best to package it up both in Docker and as an R package. I’d be curious to know what the best way to reproduce it is found to be. Working through vignettes or spinning up a Docker instance. Which is the preferred method?
- This paper is a good example of a standard social science study that is (I hope!) fully reproducible, from main analysis, to supplementary analyses and figures. - I have not yet received any external feedback w.r.t. its reproducibility, so would be interested to see if I have overlooked any gaps in the reproduction workflow that I anticipated.
The results of the individual studies (4) could be interpreted in support for the hypothesis, but the meta-analysis suggested that implicit identification was not a useful predictor overall. This conclusion is an important goalpost for future work.
We propose a simple method to retrieve optical constants from single optical transmittance measurements, in particular in the fundamental absorption region. The construction of needed envelopes is arbitrary and will depend on the user. However, the method should still be robust and deliver similar results.
It uses the drake R package that should make reproducibility of R projects much easier (just run make.R and you're done). However, it does depend on very specific package versions, which are provided by the accompanying docker image.
This paper is reproduced weekly in a docker container on continuous integration, but it is also set up to work via local installs as well. It would be interesting to see if it's reproducible with a human operator who knows nothing of the project or toolchain.