.. _Larch For XAFS Analysis (youtube.com): https://youtube.com/playlist?list=PLgNIl_xwV_vK4V6CmrsEsahNCAsjt8_Be


==================================================
Getting Started with Larch
==================================================

Larch provides several tools for working with X-ray spectroscopy data.
First, Larch provides a Python programming library that includes (or, at
least aims to include) all the functionality needed for visualizing,
processing, and analyzing X-ray Absorption and Flourescence spectroscopy
data.  Most users will start Larch has a few GUI applications, especially
``Larix``, ``GSE XRM MapViewer``, and ``LarchXRF`` for these.  In
addition, Larch includes an embedded Python-like macro language for
interacting with data that can be used either from a basic command-line
interface or as a callable service from a different programming language
(so that Athena and Artemis can use Larch instead of the older Ifeffit
library).  In fact, most of the Larch GUIs generate and run code in this
"larch macro language" so that it can be recorded for reproducible results
and to assist creating batch scripts and more complicated analysis scripts.

Because of these different levels of access available, it can be somewhat
confusing to answer the question "What is Larch?" and slightly overwhelming
for new people to know where to get started using it.

First, install Larch
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

If you are new to Larch, we recommend installing using the binary installer
for your operating system listed in the :ref:`Table of Larch Installers
<installers_table>`.  On the other hand, if you are familiar with Python
and want to use Larch as a library see :ref:`Downloading and Installation
<install-chapter>` for other ways to install Larch into your existing
Python environment.


Second, use Larix (was XAS Viewer) and/or GSE MapViewer
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

The :ref:`Larix <larix_app>` (was XAS Viewer) Graphical User Interface program
provides a complete set of tools for the visualization, processing, and
analysis of XAS data, Both XANES and and EXAFS data processing and fitting are
supported.  Some instructional videos for using Larix are at `Larch For XAFS
Analysis (youtube.com)`_.

Larix is a complete (or nearly so) replacement for Athena and Artemis
programs. There are several improvements in graphics and handling of large
data sets, and some improvements in XAFS data processing too.  In particular,
For XANES analysis, Larix includes more robust tools for peak-fitting, and
machine-learning methods such as Principal Component Analysis, Partial Least
Squares and LASSO regression.

Larix provides graphical user interface tools for running Feff to to simulate
EXAFS :math:`\chi(k)` spectra.  A graphical form for browsing 20,000 CIF
Structures from the American Mineralogist Crystal Structure Database is
included, and you can import other CIF files or structure files from DFT or
other calculations. Any of these structure files can be used to create inputs
for and run Feff (either Feff6 or Feff8l) to calculate EXAFS path
contributions and organize these results.

Larix also includes a form to build a sum-of-paths model for an EXAFS spectrum
from a set of Feff path and to help you fit a sum of Feff Paths to
experimental EXAFS :math:`\chi(k)` spectra.  This includes friendly tools to
set up EXAFS fitting parameters, working with and constraining Path
Parameters, running fits and browsing the results.

For all its processing and fitting (pre-edge peaks, PCA, Feff-fit), Larix
always saves the full of history of commands it runs as code that can be
modified or re-run in batch.  For "Feff fitting" in particular, a fitting
script can be saved for any fit and run either in the Larch macro language or
(with uncommenting of some `import` statements`) run as a standalone Python
script.  Larch can also be used as a backend for the Athena and Artemis
programs for XAFS Analysis. When you install Larch and the latest version of
Demeter, and Demeter should find and use Larch for EXAFS Analysis, replacing
the older Ifeffit library and its many limitations.

If you are a user of the GSECARS microprobe beamline or have XRF Mapping
data from a compatible beamline (XFM at NSLS-II, maybe others), you'll want
to start using the :ref:`GSE Mapviewer <mapviewer_app>` program for
reading, displaying, and working with X-ray fluorescence maps.  Much of the
documentation here discusses commands you can type in the "Larch Buffer",
available from the Mapviewer program for scripting and more detailed access
to the data in the XRF map HDF5 files.

If you are a general-purpose user or ready for more control over data
analysis for many types of data, the Larch GUI can help you browse through
the available commands and data, and provide a good starting point for
interactive, exploratory data analysis.


Third, start scripting with Larch and/or Python
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Once you've done a little bit of GUI or interactive work, you may be ready
to write scripts.  Such scripts can help you automate repeated tasks and
can build and remember more complex analyses.  The combination of the high
level commands of Larch and the interactive command-line GUI for
exploratory data analysis are a great way to get started in writing your
own scripts and building up more sophisticated programs.

The :ref:`Larix <larix_app>` application can assist you get started with this,
as it keeps a history of all commands it executes that can be saved and re-run
or modified to run in the Larch macro language or (with including the
appropriate `import` statements) as a Python program.
