Reorganize repo and add JSON file for all models (#6)

* Reorganize code files into directories

* Update readme

* Clarify that data is remote

* Add local data

* Update readme

* Correctly escape special characters

* Add missing filename

.gitignore

@@ -127,6 +127,3 @@ dmypy.json
 
 # Pyre type checker
 .pyre/
-
-# NetCDF output type checker
-*.nc

README.md

@@ -1,6 +1,8 @@
 # ArviZ `InferenceData` examples
 
-This repository contains the code used to generate ArviZ InferenceData examples that are available with `arviz.load_arviz_data`. This serves both as extra information on the models in addition to the description printed by `arviz.list_datasets` and to ease updating these example whenever necessary.
-
-Code should be added as an executable file, either `.py` or `.ipynb` in case an extended description was to be included too. The name of the file should be the same as the key passed to `load_arviz_data`
+This repository contains metadata of ArviZ InferenceData examples and the code used to generate some of the examples.
+Example models stored remotely are listed in `data_remote.json`.
+Example models stored locally are listed in `data_local.json`, and the data themselves are stored in `data/`.
+Inclusion of the code serves both as extra information on the models in the JSON files and to ease updating these examples whenever necessary.
 
+When applicable, code for a model should be added as an executable file (e.g. `.py` or `.ipynb`) to a directory in `code/` of the same name as the model.

data_local.json

@@ -0,0 +1,12 @@
+[
+    {
+        "name": "centered_eight",
+        "filename": "centered_eight.nc",
+        "description": "A centered parameterization of the eight schools model. Provided as an example of a model that NUTS has trouble fitting. Compare to `non_centered_eight`.\n\nThe eight schools model is a hierarchical model used for an analysis of the effectiveness of classes that were designed to improve students' performance on the Scholastic Aptitude Test.\n\nSee Bayesian Data Analysis (Gelman et. al.) for more details."
+    },
+    {
+        "name": "non_centered_eight",
+        "filename": "non_centered_eight.nc",
+        "description": "A non-centered parameterization of the eight schools model. This is a hierarchical model where sampling problems may be fixed by a non-centered parametrization. Compare to `centered_eight`.\n\nThe eight schools model is a hierarchical model used for an analysis of the effectiveness of classes that were designed to improve students' performance on the Scholastic Aptitude Test.\n\nSee Bayesian Data Analysis (Gelman et. al.) for more details."
+    }
+]

data_remote.json

@@ -0,0 +1,51 @@
+[
+    {
+        "name": "radon",
+        "filename": "radon_hierarchical.nc",
+        "url": "http://ndownloader.figshare.com/files/24067472",
+        "checksum": "a9b2b4adf1bf9c5728e5bdc97107e69c4fc8d8b7d213e9147233b57be8b4587b",
+        "description": "Radon is a radioactive gas that enters homes through contact points with the ground. It is a carcinogen that is the primary cause of lung cancer in non-smokers. Radon levels vary greatly from household to household.\n\nThis example uses an EPA study of radon levels in houses in Minnesota to construct a model with a hierarchy over households within a county. The model includes estimates (gamma) for contextual effects of the uranium per household.\n\nSee Gelman and Hill (2006) for details on the example, or https://docs.pymc.io/notebooks/multilevel_modeling.html by Chris Fonnesbeck for details on this implementation."
+    },
+    {
+        "name": "rugby",
+        "filename": "rugby.nc",
+        "url": "http://ndownloader.figshare.com/files/16254359",
+        "checksum": "9eecd2c6317e45b0388dd97ae6326adecf94128b5a7d15a52c9fcfac0937e2a6",
+        "description": "The Six Nations Championship is a yearly rugby competition between Italy, Ireland, Scotland, England, France and Wales. Fifteen games are played each year, representing all combinations of the six teams.\n\nThis example uses and includes results from 2014 - 2017, comprising 60 total games. It models latent parameters for each team's attack and defense, as well as a parameter for home team advantage.\n\nSee https://docs.pymc.io/notebooks/rugby_analytics.html by Peader Coyle for more details and references."
+    },
+    {
+        "name": "regression1d",
+        "filename": "regression1d.nc",
+        "url": "http://ndownloader.figshare.com/files/16254899",
+        "checksum": "909e8ffe344e196dad2730b1542881ab5729cb0977dd20ba645a532ffa427278",
+        "description": "A synthetic one dimensional linear regression dataset with latent slope, intercept, and noise (\"eps\"). One hundred data points, fit with PyMC3.\n\nTrue slope and intercept are included as deterministic variables."
+    },
+    {
+        "name": "regression10d",
+        "filename": "regression10d.nc",
+        "url": "http://ndownloader.figshare.com/files/16255736",
+        "checksum": "c6716ec7e19926ad2a52d6ae4c1d1dd5ddb747e204c0d811757c8e93fcf9f970",
+        "description": "A synthetic multi-dimensional (10 dimensions) linear regression dataset with latent weights (\"w\"), intercept, and noise (\"eps\"). Five hundred data points, fit with PyMC3.\n\nTrue weights and intercept are included as deterministic variables."
+    },
+    {
+        "name": "classification1d",
+        "filename": "classification1d.nc",
+        "url": "http://ndownloader.figshare.com/files/16256678",
+        "checksum": "1cf3806e72c14001f6864bb69d89747dcc09dd55bcbca50aba04e9939daee5a0",
+        "description": "A synthetic one dimensional logistic regression dataset with latent slope and intercept, passed into a Bernoulli random variable. One hundred data points, fit with PyMC3.\n\nTrue slope and intercept are included as deterministic variables."
+    },
+    {
+        "name": "classification10d",
+        "filename": "classification10d.nc",
+        "url": "http://ndownloader.figshare.com/files/16256681",
+        "checksum": "16c9a45e1e6e0519d573cafc4d266d761ba347e62b6f6a79030aaa8e2fde1367",
+        "description": "A synthetic multi dimensional (10 dimensions) logistic regression dataset with latent weights (\"w\") and intercept, passed into a Bernoulli random variable. Five hundred data points, fit with PyMC3.\n\nTrue weights and intercept are included as deterministic variables."
+    },
+    {
+        "name": "glycan_torsion_angles",
+        "filename": "glycan_torsion_angles.nc",
+        "url": "http://ndownloader.figshare.com/files/22882652",
+        "checksum": "4622621fe7a1d3075c18c4c34af8cc57c59eabbb3501b20c6e2d9c6c4737034c",
+        "description": "Torsion angles phi and psi are critical for determining the three dimensional structure of bio-molecules. Combinations of phi and psi torsion angles that produce clashes between atoms in the bio-molecule result in high energy, unlikely structures.\n\nThis model uses a Von Mises distribution to propose torsion angles for the structure of a glycan molecule (pdb id: 2LIQ), and a Potential to estimate the proposed structure's energy. Said Potential is bound by Boltzman's law."
+    }
+]