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.

1

Load a table

Or start from one of these

Paste instead
2

Say which column is what

Load a table first.

3

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.

4

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.

5

Solve on your levels

POST /v1/solve/pricing with the grid taken from your file and the precision from step 4.

6

Take it with you

The same run, in Python
# 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.