Conditions in limits, algebra and random models

Read geometric probability in Book Isle: one eighth of area is not one selected point out of nine

Fix a continuous uniform two-dimensional model before comparing boundary events and probabilities with nine equally likely grid points.

English translation of the Chinese original, translated on . Original publication dates, attribution, code and version notes are retained.

A continuous random location in a square and a draw from nine grid points are different models. Read region, uniformity and boundaries in Book Isle, computing the area probability of x+y≤1 before a discrete control. The app displays diagrams and available notes, not random sampling or validation of a sampling mechanism.

State the whole region and meaning of uniformity

Externally let continuous (X,Y) be uniform on the square 0≤x≤2,0≤y≤2. Equal-area regions have equal probability; drawn points or grid lines do not each receive a separate equal share. Make P1 sample space/mechanism, P2 event/areas and P3 discrete/nonuniform controls, keeping editable work and exporting coordinate-readable PDF. Check both axes reach two and bookmark P2. This is joint uniformity, not an assumption established by evenly drawn marks or colour.

Draw the triangular region, not only intercepts

Event A is X+Y≤1. Mark (1,0),(0,1), join x+y=1 and retain the origin side including its boundary, producing triangle (0,0),(1,0),(0,1). Test the origin and exclude (2,2) to select the side. Two circled intercepts alone omit the interior. Restrict shading to the square: an inequality defines an infinite half-plane, while this probability uses its intersection with the stated sample space.

Divide areas measured in the same coordinate scale

Square area is 2×2=4 and triangle area 1×1/2=1/2, so probability (1/2)/4=1/8=12.5%. The denominator is neither side length two nor three tick labels. Constant joint density 1/4 times event area gives the same result; density is not a point probability. Unequal screen-axis scaling does not change labelled mathematical areas. Do not substitute screenshot-pixel proportion for coordinate area. Record region, area and probability separately; probability itself has no square-metre unit.

continuous uniform square: 0<=x<=2, 0<=y<=2
event A: x+y<=1; area=1/2; total_area=4; P(A)=1/8
continuous: P(x+y<1)=1/8; P(x+y=1)=0
discrete grid: x,y in {0,1,2}, all 9 pairs equally likely
discrete: P(x+y<=1)=3/9; P(x+y<1)=1/9; P(x+y=1)=2/9

Zero boundary probability does not mean no equation solutions

In this continuous uniform model, x+y=1 has infinitely many points but zero two-dimensional area and thus probability zero. Replacing ≤ with < retains 1/8 by removing only that boundary. This does not mean the line has no points or the equation is logically impossible; every specified singleton also has zero probability in an ideal continuous model. Probability measure differs from the count of solutions. Keep equality, strict and nonstrict events together rather than erase the boundary and forget the model.

Use a separate model for nine grid points

If instead drawing equally from nine ordered pairs with x,y each in {0,1,2}, ≤1 includes (0,0),(1,0),(0,1), giving 1/3. Strict <1 retains only (0,0), giving 1/9; equality includes two points, giving 2/9. Positive boundary probability does not contradict the continuous case because the mechanism changed. Drawing both in one square is a comparison, not a claim that nine-point counting improves the continuous area calculation. Familiar answers do not authorize changing assumptions.

Check another threshold and a nonuniform density

In the original continuous uniform square, X+Y≤3 excludes a top-right triangle of area 1/2, giving (4−1/2)/4=7/8. For a separate nonuniform control stipulate density f(x,y)=x/4 on the same square, whose total integral is one. Integrating over 0≤x≤1,0≤y≤1−x gives P(X+Y≤1)=1/24, not 1/8. The distribution changed; the original area calculation did not fail. Without doing the integral, at least mark uniformity absent rather than use area ratio for an unspecified mechanism.

Event X+Y>1 is A’s complement and also has probability 7/8, but differs from X+Y≤3. The origin belongs only to the latter; (2,2) only to the former. Equal probability does not identify an event, just as equal-area regions may occupy different locations. Retain conditions with the fraction rather than save the result alone.

Inspect the sampling mechanism before the shading

For 1/3, check accidental use of the nine-point model; for 1/4, check missing triangle factor one-half. Unchanged probability after ≤ becomes < belongs to the continuous case, not the discrete table. Retaining 1/8 for the nonuniform example treats variable density as constant. Repair one note-copy assumption and return to square limits/mechanism. Scattered-looking points do not prove uniformity; a small empirical frequency need not equal theory exactly. Leave observations blank if no sampling occurred.

Reopen and deliver answers for both models

Reopen P2, state continuous uniform sampling, then triangle area 1/2 and probability 1/8. On P3 check discrete 1/3, strict 1/9 and boundary 2/9, explaining why the continuous boundary is zero. Deliver sample space, mechanism, event region and density control rather than shading alone. Mark unperformed bookmark/highlight/reopen actions pending; claim no Book Isle sampling, integration or device evidence of real-world uniformity.

Read probability examples in Book Isle: separate mutually exclusive and independent events with four outcomes; Book Isle conditional probability: changing the given condition changes the denominator.

Method references: UT Austin Žitković Joint Distributions §5.3. Diagrams, report data, counterexamples and ledgers are original exercises.

Version checked on 18 November 2026: the Apple public lookup lists Book Isle 1.1.4, requiring iOS / iPadOS 15.6 or later; see the official App Store 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.