Bring your own data
Load a table of choices somebody actually made — prices set, bids submitted, offers made — and this page fits the precision those choices imply, then solves the problem on the levels your file contains. Five steps, two API calls, and a script at the end that reproduces both.
Load a table
Or start from one of these
Paste instead
Say which column is what
Load a table first.
Name the problem
The levels come from your file. These are the numbers no table can supply, and they are what turns a column of choices into a payoff a precision can be fitted against.
Estimate what the data can pin down
POST /v1/fit takes the payoff table above and your observed counts and returns the precision that best explains those choices, with an interval that names the method that produced it. Where the likelihood is flat it returns no number at all, because a refusal is a bound and not an estimate.
Solve on your levels
POST /v1/solve/pricing with the grid taken from your file and the precision from step 4.
Take it with you
# Load a table and map its columns to see the script.
pip install strataq. This script does the fit and the solve locally, with no service in the middle, and returns the same numbers. The API console has the HTTP form of both calls.