Manuscript XXII-C@1 · XXII-C.tex · section · b9a05a24

What convergence of process data can mean

Continuum, refinement, and evidence discipline › What convergence of process data can mean · lines 1168–1208

XXII-C@1#eq:process-distance, XXII-C@1#eq:finite-family-convergence, XXII-C@1#eq:uniform-family-convergence

Back to manuscript

For two processes on the same frozen interface define

This is an operational probability-comparison quantity for the specified one-use tests. It is not a diamond norm on an unspecified global channel.

A family of approximating data could be compared by, for example,

or by the stronger uniform statement

A future argument must name its topology and prove the convergence it uses. Neither statement is built into Definition def:class.

Even convergence of all strengths does not automatically yield well-defined maps between the component quotients. A positive limiting strength is eventually witnessed by positive approximating strengths, but a zero limiting strength may have positive approximating strengths at every scale, as in Proposition prop:strength-stability. Component splitting or merging therefore requires its own exact support or compatibility theorem.

Displayed formulas in this passage

  • XXII-C@1#eq:process-distance · line 1171

  • XXII-C@1#eq:finite-family-convergence · line 1185

  • XXII-C@1#eq:uniform-family-convergence · line 1192