Stat-Ease 360 and Python demo

This is a collection of a few Python programs that interact with Stat-Ease 360, initially demonstrated at the 2021 DOE Summit by Hank Anderson. The slides from that talk are included in this repository as well.

In order to use the data from this repository, you should install the se360demo Python package by running pip install se360demo.

Multi-response Graph

This uses the examples\multiresponse-graph.py script. Open Stat-Ease 360 and load this script from the Script Window, then click File -> Run.

It will retrieve the two Analysis objects from SE360 and make predictions for both of them while varying the A factor. These predictions are plotted on the same Matplotlib plot.

Streamflow

This example uses the examples\streamflow.py script.

It demonstrates fetching data from a cloud API and inserting it into Stat-Ease 360. It has a flag in the script to toggle between live data and a csv file.

The data retrieval and manipulation code are in se360demo\__init__.py. To use the live data you'll need to set an environment variable called NOAA_TOKEN with a token generated from https://www.ncdc.noaa.gov/cdo-web/token.

Delivery Time

This example uses the examples\delivery-time.py script.

This is the Delivery Time example from chapter 11.2.3 of Montgomery, Peck and Vining (2012) on "data splitting" or cross validation. The data are first segmented and evaluated according to the textbook. The data are then split into 4 segments for a k-fold cross validation. Each split is evaluated and the results are plotted in a bar chart.

Simulated Cross Validation

This example uses the examples\sim-cv.py script.

This uses two data sets that both have simulated responses. The script compares two analyses of each response using cross validation. One analysis is done using Ordinary Least Squares and the other using a Gaussian Process Model.

Streamflow Cross Validation

This example uses the examples\streamflow-cv.py script.

This does a time-based cross validation on the streamflow data. It divides the data up into years for one cross validation, then does another cross validation based on the month of the year. The results are output to the console.