Methods

A page in this section sets out what Tidemark computes from the data, the pass schedule included, and why each step is built the way it is, for a reader who wants the reasoning behind the code.

The Analysis Graph takes each part of that in turn:

Section	What It Covers
Declaring a Step	How a step is written, and how the order among them is read out of the functions
Anchors and Targets	Which gauges the first steps hold as anchors and which as targets
Each Step's Key	What a step's result is kept under, and which steps compute on every run
Every Setting Declared Once	Where an analytic choice is written down, and how to change one
Each Run's Record	What each run records beside its results, and where to browse the runs recorded so far
Checking Each Table	How a table breaking its schema stops the run where it was produced

The Pass Schedule then sets out the passes covering each gauge, each gauge's place across the swath, the looks due there and those the fetched passes hold, the gauges one pass number alone covers, and the columns of each table the package ships for it.

From Samples to One Height sets out which samples around a gauge count toward a pass, which level of product each pass reads, and how each pass reduces to one height beside its uncertainty.

Reproduce then takes a reader from an empty directory to a machine that fetches the data and runs the comparison, and sets out what a change passes before it reaches main.