pyDAMPF: a Python package for modeling mechanical and electrostatic properties of hygroscopic materials under interaction with a nanoprobe

pyDAMPF is a tool oriented to the Atomic Force Microscopy (AFM) community, which allows the simulation of the physical properties of materials under variable relative humidity (RH).

The computing engine is written in Fortran in order to reuse physics code and wrapped to Python to interoperate with high-level packages. We also introduce an in-house multi-thread approach of the pyDAMPF code, which compares for various computing architectures (PC, Google Colab and a HPC facility) very favorable in comparison to a former AFM simulator.

Users can use the existing cantilever database to perform their simulations or add-up further models. There are three ways of running pyDAMPF:

• serial (uses a single thread)
• multi-thread (unix based computer)
• multi-thread (SLURM based computer)

Documentation and learning material are also available in the form of:

• An introduction to pyDAMPF in the context of evaluating maximum exerted forces by a cantilever Tip under variable environments is in our SciPy 2022 paper.

• You may also try pyDAMPF online for free on a Google Colab notebook.

• A short tutorial on how to use pyDAMPF (YouTube).

Installation

pyDAMPF is itself largely Python and Fortran but depends on numpy, matplotlib, and scipy

You should be able to download pyDAMPF by doing::

$ git clone https://github.com/willymenacho/pyDAMPF.git

pyDAMPF has a numerical kernel in fortran so it is necessary to install the correct version.

$ sudo apt-get update

$ sudo apt-get install gfortran-7-multilib

Compilation with f2py: This step is only required once,and depends on the computer architecture, by using f2py with the file pyDAMPF.f90 within the folder EXECUTE_pyDAMPF, the code for this reads

$ f2py -c --fcompiler=gnu95 pyDAMPF.f90 -m mypyDAMPF

$ cp *.so ~/pyDAMPF/EXECUTE_pyDAMPF/pyDAMPF_BASE/nrun/

$ cp *.so ~/pyDAMPF/EXECUTE_pyDAMPF/pyDAMPF_BASE/nrun/runa

RUN pyDAPMF

Interactively from a Jupyter Notebook

Once you have followed the installation steps, pyDAMPF is ready to run. In our latest version, we have a user friendly interface, you can see a short tutorial in pyDAMPF (YouTube).

$ jupyter notebook interfaz_v1.ipynb

Also, an interactive server will be soon available through cloud computing now on Binder.

Local excecution with python

Once we have obtained the numerical code as Python modules we generate the tempall.txt file which contains all the necessary parameters and variables for the pyDAMPF execution.

$ python3 inputs_processor.py

pyDAMPF excecution modes

We need to choose the execution mode which can be serial or parallel. Whereby parallel refers within this first version of the code to multi-threading capabilities only.

Serial method: Our in-house development creates an individual folder for each simulation case,which can be executed in one thread.

$ python3 serial_method.py

Parallel method:

This method comprises two parts:

First a function which takes care of the bookkeeping of cases and folders:

$ python3 parallel_method.py <number of threads>

The second part of the parallel method will execute pyDAMPF, which contains at the same time two scripts. One for executing pyDAMPF in a common UNIX based desktop or laptop.

$ python3 job_parallel_computer.py <number of threads>

While the second is a python script which generated SLURM code to launch jobs in HPC facilities

$ python3 job_parallel_cluster.py <number of threads>

Analysis

Once the pyDAMPF simulation is finished, pyDAMPF has two ways of analyzing the data.

The graphical analysis:

$ python3 Graphical_analysis.py

The quantitative analysis:

$ python3 Quantitative_analysis.py

Alternatively we offer for both cases an interactive environment in jupyter notebook.

$ pip install tabloo

$ jupyter notebook Graphical_analysis.ipynb

$ jupyter notebook Quantitative_analysis.ipynb

Example

To relate to the use of pyDAMPF you can access Google Colab notebook