pySpecData 0.9.5.1.1

Features

• Labeled axes allow one to manipulate datasets (potentially with different dimensions) without having to explicitly keep track of what the different dimensions correspond to. Code becomes more legible. Also, tiling, direct product, and gridding functions become obsolete.

• Fourier transformation with automatic manipulation of axes.

• Automatic error propagation.

• Commands like plot(data) will generate a plot with automatically labeled axes, errors, and units. All of this information is also written to HDF5 when the data is saved.

• Simplified curve fitting that takes advantage of labeled axes and Python’s symbolic algebra package (sympy).

• The code is written so that it can be integrated into a nicely formatted PDF lab notebook.

  • The same code can be run on the command line (to generate pop-up plot windows) and embedded into a LaTeX document.

  • Extension to other output formats, such as HTML or markdown, should be relatively straightforward.

• In a multimedia environment like jupyter, you don’t need a separate plot command. The code can automatically choose a plotting style appropriate to the code (eventually, the general preferences for this can just be configured at the beginning of the jupyter notebook).

More detailed web documentation will be coming soon.

NMR/ESR specific

Because it was written primarily for NMR and ESR data, it also includes:

• Routines for reading commercial raw data (e.g. Bruker, Kea) into nddata objects with all relevant information.

• The object-oriented features make it much easier to process raw phase-cycled data and to simultaneously view multiple (potentially interfering) coherence pathways.

• Contains functions for baseline correction, peak integration, etc.

• (Not yet in packaged version) A basic compiled routine for propagating density matrices that can be used to predict the response to shaped pulses.

Version Notes

Note that the current version is intended just for collaborators, etc. (Though, if you do really want to use it for interesting science, we are happy to work with you to make it work for your purposes.) A public-use version 1.0.0, to be accompanied by useful demonstrations, is planned within a year. (Note that the email currently linked to the PyPI account is infrequently checked –if you have interest in this software, please find J. Franck’s website and contact by that email.)

Installation Notes

Highly Recommended: Install the following packages using a good package-management system (conda or linux package manager), rather than relying on pip or setuptools to install them:

• numpy

• scipy

• sympy

• pyqt

• pytables

• matplotlib

• h5py

For example, on Windows with Anaconda 2.7. – just run conda install numpy scipy sympy pyqt pytables matplotlib h5py.

(If you don’t install these packages with your system pip will try to install them, and there is a good chance it will fail – it’s known not to work great with several of these; setuptools should error out and tell you to install the packages.)

mayavi: Mayavi can be used (and gives very nice graphics), but frequently lags behind common Python distros. Therefore, this package was written so that it doesn’t depend on mayavi. Rather, you can just import mayavi.mlab and pass it to any figure list that you initialize: figlist_var(mlab = mayavi.mlab)

If you are installing from github (or generally using setuptools – i.e. python setup.py install or python setup.py develop), you will need some additional packages to enable compilation:

• libpython

• unxutils

• mingw

The last two are specific to Windows, and provide things like the gcc and gfortran compiler, as well as make.

(Once these are installed, to install from github, just git clone https://github.com/jmfranck/pyspecdata.git then move to the directory where setup.py lives, and do python setup_paramset.py develop followed by python setup.py develop)

Quick-Start

To get started with this code:

1. Install a good Python 2.7 distribution

   • On Windows or MacOS: Anaconda 2.7. When installing select “install for all users.”

2. Install libraries that pyspecdata depends on. (If you’re interested in why you need to do this first, see installation notes below.)

   • On Windows or MacOS: in the Anaconda Prompt, run conda install numpy scipy sympy pyqt pytables matplotlib h5py.

   • On Linux, just use your package manager (aptitude, yum, etc.) to install these libraries.

3. Install pyspecdata: pip install pyspecdata

4. Set up directories – create a file in your home directory called _pyspecdata (Windows – note the underscore) or .pyspecdata (Mac or Linux). Here is an example – you can copy and paste it as a starting point:

   [General]
   data_directory = c:/Users/yourusername/exp_data
   notebook_directory = c:/Users/yourusername/notebook

   Note that any backslashes are substituted with forward slashes. Also note that you will need to change the directories to refer to real directories that already exist or that you create on your hard drive (see below). Note that on Windows, you can use notebook, etc. to create this file, but it cannot have a .txt, etc. extension.

   • Where is my “home directory”? (Where do I put the _pyspecdata file?)

     • On Windows, your home directory is likely something like C:\Users\yourusername. You can access your home directory by opening any file folder window, and starting to type your name in the address bar – it’s the first folder that shows up underneath.

     • On MacOS and Linux, it’s the directory indicated by ~. On Linux, this typically expands to /home/yourusername.

   • What are these directories? → You can either create them or point to existing directories.

     • data_directory must be set. It is a directory, anywhere on the hard drive, where you store all your raw experimental data. It must contain at least one subdirectory – each subdirectory stores different “experiment types,” typically acquired on different instruments (e.g. you might have subdirectories named 400MHz_NMR, 500MHz_NMR, 95GHz_ESR, and Xband_ESR).

       • Data is assumed to be unpacked (i.e. as it is on the spectrometer – not in .zip or .tgz files)

       • If you’re setting up a lab, you might want to separately sync each different experiment type folders using seafile.

         Or you can sync the whole data directory with dropbox.

     • If set, the notebook_directory is intended to contain latex files with embedded python code, as well as some processed output.

   • Do not use quotes to surround the directory name. Even if it contains spaces, do not use quotes, and do not escape spaces with backslashes.

   • Note that on Windows, your desktop folder is typically in C:\Users\yourusername\Desktop

   • Why do I need to do this?

     • Setting this configuration allows you to move code between different computers (e.g. a spectrometer computer, a desktop, and a laptop), and re-use the same code, even though the locations of the files are changing. This should work even across different operating systems.

     • It specifically enables functions like find_file(...), get_datadir(...), etc. that can search the data directory for a file name matching some basic criteria. You should always use these to load your data, and never use the absolute path.

     • The GUI tool that will allow you to set up _pyspecdata by pointing and clicking has not yet been set up.