# Read Cauchy–Schwarz in Book Isle: computing a bound still leaves an equality condition

Language: en
Canonical: [Read Cauchy–Schwarz in Book Isle: computing a bound still leaves an equality condition](<https://allluckly.cn/en/guides/nasbook-cauchy-schwarz-bound-equality-zero-vectors-two-dimensional/>)
中文：[中文 HTML](<https://allluckly.cn/guides/nasbook-cauchy-schwarz-bound-equality-zero-vectors-two-dimensional/>)
English: [English HTML](<https://allluckly.cn/en/guides/nasbook-cauchy-schwarz-bound-equality-zero-vectors-two-dimensional/>)
Published: 2026-11-18T00:00:00+08:00
Modified: 2026-11-18T10:24:20+08:00
Author: Bison

Check four below five, a squared-determinant gap, positive/negative multiples and zero-vector boundaries in real two-dimensional vectors.

English translation of the [Chinese original](<https://allluckly.cn/guides/nasbook-cauchy-schwarz-bound-equality-zero-vectors-two-dimensional/>), translated on 2026-11-18. Original publication dates, attribution, code and version notes are retained.

A numeric right-hand bound does not mean every input attains it. Read Cauchy–Schwarz in Book Isle with separate dot-product, length-product and equality columns, then return to a constrained maximum. Compute externally; matching highlight colours do not prove proportional vectors.

## Fix real two-dimensional coordinates and Euclidean length

Externally set u=(1,2), v=(2,1) with dimensionless real coordinates, usual dot product and Euclidean norm. Use P1 definitions, P2 inequality/exact checks and P3 equality/constraint controls, keeping editable work and exporting readable PDF/text. Verify coordinate order and bookmark P2; highlight conditions only with appropriate controls, otherwise record externally. This is not a complex-inner-product problem, and the unconjugated real formula is not automatically valid for arbitrary complex coordinates. Visual proximity of arrows does not establish parallelism.

## Compute dot product and product of lengths separately

The dot product is 1×2+2×1=4. Both squared lengths are five, so each norm is √5 and their product five, not √10. Cauchy–Schwarz bounds the absolute dot product by that product; four is strictly below five here. Label 4,√5,√5,5 by role, not as angle degrees or a sum of lengths. In the squared version the sides are sixteen and twenty-five; square both sides consistently rather than only one.

```text
u=(1,2); v=(2,1)
u dot v=4; norm(u)^2=5; norm(v)^2=5
abs(u dot v)=4<5=norm(u)*norm(v)
5*5-4^2=9=(1*1-2*2)^2
w=3*u: dot=15, norm_product=15
w=-3*u: dot=-15, absolute_dot=15, norm_product=15
w=(0,0): both sides=0; no angle quotient
```

## Check the gap with a two-dimensional square identity

For real (a,b),(c,d), expansion gives (a²+b²)(c²+d²)−(ac+bd)²=(ad−bc)². The square is nonnegative, independently proving the two-dimensional inequality. Here ad−bc=1×1−2×2=−3, whose square nine equals twenty-five minus sixteen. Retain the minus sign inside the square rather than call −3 the inequality gap. This algebra does not depend on drawing precision or measured screen angles.

## Check equality for positive and negative multiples

With w=3u=(3,6), dot and norm product are fifteen, attaining equality. With w=−3u=(−3,−6), dot is negative fifteen but absolute dot still fifteen, again equality. Proportional nonzero real vectors include opposite directions for the absolute-value bound. In the upper-bound statement without absolute value u·w≤norm product, the opposite direction instead gives the negative lower endpoint. Separate absolute-value equality from a positive maximum; a negative dot product does not rule out equality in Cauchy–Schwarz.

## Zero vectors attain equality but cannot divide a norm

For w=(0,0), both dot and norm product are zero, so equality holds. Do not divide by its norm to form a scalar ratio or assign an angle from 0/0. State the complete condition: either vector zero gives equality; with both nonzero, equality holds exactly for proportional vectors. The square identity includes zero cases. Original u/v have equal norms but determinant −3≠0, so are not proportional. Equal length alone is insufficient; keep equal norm and a common direction line separate.

Doubling original v gives dot eight and norm product ten, still strict. Multiplication by zero yields equality at zero without proving original proportionality. This degenerate transformation loses information about the original equality case. Recheck zero status before extending a nonzero-scaling observation to a zero multiplier.

## Carry equality into a constrained maximum

For real x,y with x²+y²=5, target x+2y is u·(x,y). Both norms are √5, so absolute value is at most five and the positive maximum at most five. Candidate (1,2) satisfies the constraint and attains five, proving attainability; (−1,−2) gives minimum negative five. Original (2,1) is feasible but yields four, not the maximum. Do not stop at the bound or use (3,6) as a maximizing candidate: squared norm forty-five violates this constraint.

With x²+y²\<5 instead, the target is below five: five is an approachable bound, unattained, and (1,2) is infeasible. Along t(1,2),0\<t\<1, values 5t approach five without attaining it. With ≤5 the boundary is feasible again. Check strictness together with existence of a maximum rather than reuse maximum five merely because the bound number stayed unchanged.

## Troubleshoot absolute value, norms and premises

For √10 on the right, inspect adding squared lengths versus multiplying norms. A gap −3 misses the final square. For a rejected opposite-direction equality, check absolute-value bound versus positive upper endpoint. At zero, stop angle/ratio division and substitute directly. Correct one note-copy item, returning to assumptions. Complex coordinates or a changed inner product require their own definition; an equality search hit does not test feasibility of a constrained candidate.

## Reopen with bound and attainability checks

Reopen P2 without answers, give four and five and derive the gap nine from the square identity. Explain positive multiple, negative multiple and zero on P3. Finish with maximum five and candidate (1,2), checking x²+y²=5. Deliver values, proof, equality conditions and an infeasible candidate, not a standalone inequality. Mark unperformed bookmark/highlight/reopen actions pending; no automatic Book Isle optimization, parallelism detection or device precision test is claimed.

## Related tutorials and sources

[Book Isle dot products: check nonzero vectors before claiming a right angle](<https://allluckly.cn/en/guides/nasbook-dot-product-orthogonality-nonzero-angle-boundary/>); [Read eigenvectors in Book Isle: scalar-multiple output still requires nonzero input](<https://allluckly.cn/en/guides/nasbook-eigenvector-scalar-multiple-nonzero-two-by-two/>).

Method references: [MIT Michael Andrews 18.100B/C solutions: Cauchy–Schwarz equality](<https://math.mit.edu/classes/18.100B/fall_2011/sol2.pdf>). Diagrams, report data, counterexamples and ledgers are original exercises.

Version checked on 18 November 2026: the [Apple public lookup](<https://itunes.apple.com/lookup?id=6766198898&country=cn&entity=software&timestamp=20261118>) lists Book Isle 1.1.4, requiring iOS / iPadOS 15.6 or later; see the [official App Store](<https://apps.apple.com/cn/app/id6766198898>) for published capabilities. Use self-made PDFs, EPUBs or clear text materials; bookmarks, highlights and notes depend on format and visible controls, without automatic solving or literary analysis. External preparation and paper predictions do not constitute device observations. Leave unperformed checks blank.

