Frequently Asked Questions¶

Is T_ij going from i to j or j to i?¶

This is a confusing topic and depending on the field, other conventions are used. Here, we rely on the definition that $T_{ij}$ is the probability that we jump from $i\to j$. This means—due to probability conversation—that we have a row-normalized transition matrix $1 = \sum_j T_{ij}$ and that we need to estimate the left-handed eigenvectors for estimating the stationary distribution or implied timescales.

Is Python 3.7 supported?¶

Starting from version 1.1.x Python 3.7 supported will be dropped, due to its EOL status. If support is needed, please use the latest 1.0.x version.

How is the performance compared to PyEmma?¶

This depends heavily on the task, but in general it should be comparable or even faster, see Benchmark.

Is there a command line interface?¶

Yes, indeed. Some useful plots (e.g. CK-test) and tools can be used directly from the command line. Please check out the short tutorial CLI.

Is there a shell completion¶

Using the bash, zsh or fish shell click provides an easy way to provide shell completion, checkout the docs. In the case of bash you need to add following line to your ~/.bashrc

eval "$(_MSMHELPER_COMPLETE=bash_source msmhelper)"

In general one can call the module directly by its entry point $ msmhelper or by calling the module $ python -m msmhelper. For enabling the shell completion, the entry point needs to be used.

How do I use the Hummer-Szabo projection?¶

Following the tutorials, you find in Hummer-Szabo projection a detailed explanation. To make it short, simply create a state trajectory traj = mh.LumpedStateTraj(macrotrajs, microtrajs) and pass this object to the analysis methods.

I get negative values for the Hummer-Szabo projection?¶

This is sadly not a bug, but a limitation of the projection formalism. This should occur only for bad lumping and with values close to zero. For a detailed description of this issue please take a look at the original publication, (10.1021/jp508375q). To avoid it, you can use the flag positive=True while initialization to enforce $T_{ij} \ge 0$.

Feature X is missing¶

First, if you are looking for a feature complete package, I would recommand you to take a look at pyemma and msmbuilder. If you believe that a crucial functionality/method is missing, feel free to open an issue and describe the missing functionality and why it should be added. Alternatively, you can implement it yourself and create a PR to add it to this package, see contributing guide.

I found a bug. What to do next?¶

If you find a bug in this package, it is very kind of you to open an issue/bug report. This allows us to identify and fix the problem, thus improving the overall quality of the software for all users. By providing a clear and concise description of the problem, including steps to reproduce it, and relevant information such as device, operating system, and software version, you will help us resolve the problem quickly and effectively. Submitting a bug report is a valuable contribution to the software and its community, and is greatly appreciated by the development team.

Is it possible to build the documentation for offline use?¶

Yes, for sure. You can compile the documentation on your local machine by executing the following commands:

# install all additional dependencies
python -m pip install msmhelper[docs]
# build the docs inside the site directory
python -m mkdocs build