If a flat file of food and ingredients with the most common nutrients is all you want, we've made that available at data.world
If you want a denormalized, integrated MongoDB database of the USDA's National Nutrient database, then please see below.
Import USDA National Nutrient Database ASCII files in to a MongoDB collection.
We developed this small python against SR27. Those data files are available from the download link at http://www.ars.usda.gov/Services/docs.htm?docid=8964
Some of the bulk operations we use weren't available until PyMongo version
3.0. Ubuntu packages a different version. As a result, we have supplied a
setup.sh
script for setting up a virtualenv, and a run.sh
to use it. If
you're already familiar with using virtualenv's then you can do it all yourself:
$ cd nndb-import
$ virtualenv venv
$ . venv/bin/activate
$ pip install -r requirements.txt
$ ./nndb_import.py
./nndb_import.py
requires a specified directory containing the ASCII files
linked above. Run ./nndb_import.py --help
for details.
There is a small attempt at a GA optimization approach for recommended nutritional intake based on an attempt at a fitness function tailored to my goals. It's there for historical purposes.
I won't keep you in suspense. The best recommendations generated were either nonsensical (3 pounds of celery seed?!) or a version of Michael Pollan's advice:
Eat food. Not too much. Mostly plants.
This is a Python 3 app - we're not supporting Python 2.7