Keep the labels when operating on matrices

Try

Two classrooms store markers and notebooks. Let rows be East, West, in that order, and columns be markers, notebooks. Their starting stock is S=[[10,6],[8,12]] items. A delivery D=[[4,3],[2,5]] arrives before use U=[[9,4],[7,8]]. What remains?

Compute S+D−U entry by entry to obtain [[5,5],[3,9]]. The first row means five markers and five notebooks in East. The second means three markers and nine notebooks in West. A matrix is a rectangular array of numbers, but this contextual matrix also carries labels and units.

Predict

Predict whether the same arithmetic would be meaningful if the delivery table listed notebooks first and markers second. The dimensions would still be 2×2, but adding corresponding positions without reordering would mix different objects.

Predict what a negative remaining count would mean. Matrix subtraction is mathematically defined even when it yields negative numbers. In this stock model, however, a negative count signals that the stated use could not have occurred from the available stock. Algebraic validity and contextual feasibility are separate checks.

Experiment

Rebuild the calculation from this labeled ledger. Each cell is an equation for one kind of object in one location.

Location and item	Start	Delivery	Use	Remaining
East markers	10	4	9	5
East notebooks	6	3	4	5
West markers	8	2	7	3
West notebooks	12	5	8	9

Change only East marker use from 9 to 15. The new remaining entry is −1. Explain the failure in words rather than replacing it with zero: silently clamping would conceal a shortage of one marker. If deliveries arrived after use, even the original nonnegative final matrix might be insufficient to establish a feasible schedule; timing is an assumption.

Next, reverse the delivery’s column order and then restore it before adding. Record the reordered entries, not just a note saying that labels matter.

Understand

A matrix with m rows and n columns has dimensions m×n. Entry (i,j) lies in row i, column j. Addition and subtraction require the same dimensions and act on corresponding entries. For contextual data, those entries must also refer to compatible labels and units. A 2×3 table cannot be added to a 3×2 table by inventing a correspondence, and a count cannot be subtracted from a dollar cost without a stated conversion.

For a scalar c, the matrix cA has entry c times each entry of A. If a payoff matrix is P=[[3,−2],[1,0]] dollars per play, then 2P=[[6,−4],[2,0]] can describe two repeats of the same choice pair under an unchanged payoff rule. Positive entries are gains, negative entries losses and zero break-even. Changing choices between plays would require adding the appropriate outcomes rather than simply doubling one entry.

A 10% price increase is multiplication by 1.1, not addition of 0.1 to every price. A percentage is relative to each original entry. A dimensionless multiplier preserves the price unit; multiplying a per-item price by a number of identical items changes the resulting interpretation to total cost. As with velocity and time, the scalar’s meaning matters.

All these operations are entrywise. Matrix multiplication is a different operation, developed next. Multiplying entries in the same positions is not the standard matrix product.

Apply

Design your own two-row, three-column table, state the row/column order and units, and propose one meaningful addition and one meaningful scalar multiplication. Then give an operation that fails because of labels or units even though the shapes match.

In practice, write both the resulting array and an interpretation of at least one entry. A correct unlabeled array can show arithmetic, but not whether you understood the data model. If an entry is infeasible, preserve it as evidence of the problem and explain the assumption that failed.

Continue

This lesson supplies concrete data, scaling and addition/subtraction actions for N-VM.C.6–8. The examples use small invented counts, not a real stock-management system; the learner’s task is to reason about the stated model.

Continue to row-by-column multiplication, its dimension requirements, and the distinction between zero and identity matrices. Some familiar number laws survive, while multiplication order becomes important.