A collection of tools written in Python 3 to wrangle data in Axiell EMu for Mineral Sciences at NMNH.

The minsci package exists on PyPI but is badly out-of-date, so it's better to install directly from github. The recommended installation process uses conda and is as follows:

conda create -n nmnh_minsci
conda activate nmnh_minsci

git clone https://github.com/adamancer/nmnh_ms_tools
cd nmnh_ms_tools
conda env update -f environment.yml -n nmnh_minsci
pip install .


cd ..
git clone https://github.com/adamancer/minsci
cd minsci
conda env update -f environment.yml -n nmnh_minsci
pip install .

Once you've finished installing the package, there are a few clean up steps that you can do to better tailor things to your EMu:

• Copy your institution's schema.pl file to minsci/xmu/files. This allows the script to validate paths when reading and writing data.
• Define grids in minsci/xmu/files/tables. This allows the script to verify that all columns in a given grid have the right number of rows. The default grids are NMNH-specific and should be deleted and replaced. In particular, it's a good idea to define any grids you'll be writing to.

Most common operations involved subclassing XMu to read an export file. A very basic framework for this operation is:

from minsci import xmu

class XMu(xmu.XMu):

    def __init__(self, *args, **kwargs):
        super(XMu, self).__init__(*args, **kwargs)
        self.records = {}


    def iterate(self, element):
        rec = self.parse(element)
        # Do stuff...

xmudata = XMu("xmldata.xml")
xmudata.fast_iter(report=1000)

Examples of common operations, including sample EMu export files, are provided in https://github.com/adamancer/minsci/tree/master/minsci/examples.

The minsci module also includes a command-line utility that can be used to download specimen data from the NMNH Geology Collections Data Portal. Downloads are CSV files using the Simple Darwin Core data standard.

Here is an example download command:

minsci download -classification basalt -state hawaii

Use the -h flag to see the available options:

minsci download -h