Version 0.1.2 (Current Develop)

Rename project to metamodeling framework
Allow results_file to try and find the results in the post_process/{metamodel['name']}/simulation_results.csv}
Move downloading of simulations to Python. In custom script call from metamodeling.post_process import OpenStudioServerAPI
Remove Ruby and add post processing python code. User will need to write their own post processing scripts.
Remove the ANOVA plots in LinearModel due to missing statsmodel package.
Remove the "results_directory" key in the metamodel definition file. This was no longer used.
Add random to the validation_id option in the metamodel definition
Remove some hard coded validation checks such as calculating the total HVAC energy
Do not remove inspection results from the data folder when running build.
Rename ETSModel to Metamodel. Metamodels point to a Metamodel (not ETSModel)
Update SciKit Learn to 0.22.2, Update NumPy to 1.18.3, and other dependencies.
Use pipelines for Random Forest and enable categorical variables
Fix EPW Day of week calculation. Add Day of Week (%A) and Day of Week Int (%w).
Add seaborn-qqplot to requirements
Fix bug passing downsampling value to validation (users needs to ensure there is a downsample key in the metamodel.json file)
Remove the renaming of ETS and district energy variables. These now must be renamed in the postprocessing of the simulation results.
Updated to Seaborn 0.11, made sns calls use x, y
Move to GitHub Actions for CI

Version 0.1.1

Adds an "inspect" method to the rom-runner (rom.py) CLI
Remove example files
Updates the dependency (setup) file to ensure correct installation of dependencies
Use Python 3's super syntax
Move example runs to a unittest
Add Python coverage
Move Ruby post processing script to 'post_process' folder
Add results_file to the metamodel.json config file to point to the CSV that needs to be loaded
Moved unit tests to root directory
Moved post-processing script (which is in Ruby) from the data directory to the post_process directory