Skip to the knowledge library

explanation

How much information is enough to reconstruct a geometry?

Enough means that no two states with the same observation demand different answers to the question you are asking.

Exact deterministic reconstruction and finite coding results under the stated conditions; approximate and noisy reconstruction need further criteria.

Name the answer you need

Let E(w) be the available observation of an allowed state w, and g(w) the result required. Exact reconstruction is possible precisely when any states with the same E also have the same g. If an observation leaves two different required answers possible, a decoder cannot choose the correct one from that observation alone.

There are three honest ways forward: acquire a separating observation, justify a constraint that rules out an alternative, or ask a less detailed question. Redrawing the same information with more decoration cannot supply the missing distinction.

E(w₁)=E(w₂) ⇒ g(w₁)=g(w₂)

A triangle shows the missing distinction

Place A=(0,0) and B=(d,0). Knowing the distances AC=a and BC=b gives x=(a²+d²−b²)/(2d), then y²=a²−x². For a nondegenerate triangle, positive and negative y remain possible.

With d=4, a=3 and b=5, C can be (0,3) or (0,−3). The same three side lengths fix the triangle’s shape including its reflected alternative. An observer asking for its orientation also needs to know which side of the directed baseline contains C.

A finite count answers a finite task

If a noiseless fixed-length binary message must distinguish N possible answers, it needs at least ceiling(log₂N) bits. This bound assumes a shared codebook and context. It counts required distinctions, not a universal number of sensors, dimensions or depths.

For measurements with error, specify tolerance and conditioning. Exact coordinates drawn from an unrestricted continuum cannot all be encoded in one fixed finite number of bits.

Keep the measurement task distinct from the observer

Here, observation names the stated measurement, sampling or reconstruction operation. The operator’s mathematical conditions do not alone supply a Mind. In the current Y terminology, the observer is the continuing internally differentiated Mind that can relate its own patterns to outside relations and use that relation for intentional correction.

Read observation and the observer.

Questions to carry forward

Is there a universal percentage of a shape that reveals the whole?

No. Known symmetry and constraints can make little data sufficient. Unknown independent freedoms require more. The question and the admissible family decide what is enough.

Where this belongs in the journey

Sources and further reading