Partitioning an Interval

Why Partition an Interval?

A finite accumulation is built by adding contributions from parts of a larger interval. Before those contributions can be organized, the interval must be divided clearly: where does each part begin, where does it end, and which contribution belongs to it?

For example, a time interval from \(0\) to \(12\) seconds might be divided at times \(0\), \(3\), \(6\), \(9\), and \(12\) seconds. Those five times define four consecutive subintervals. Each contribution in a finite sum must correspond to one of those four subintervals, in the same order.

Keeping the endpoints and intervals distinct prevents a basic but common error: confusing the number of listed endpoints with the number of pieces they create.

Definition of a Partition

The endpoints must be listed in their natural order along the interval. Each subinterval runs from one endpoint to the next; a subinterval does not skip over an intermediate endpoint. Consecutive closed subintervals share a boundary endpoint, but their interiors do not overlap, and together they cover the original interval.

The index \(n\) counts the subintervals. There are \(n+1\) endpoints: the first endpoint starts the first subinterval, and each subsequent endpoint finishes one subinterval and starts the next, except that the final endpoint only finishes the last one.

This notation works for time, distance, or any other quantity measured along an interval. For time, the endpoints might be moments in seconds. For distance, they might be locations along a measured route. The roles of the endpoints and intervals remain the same.

Matching Contributions to Intervals

Suppose the contribution associated with the \(i\)th subinterval is denoted by \(C_i\). The index \(i\) starts at \(1\), because the first subinterval is the one from \(t_0\) to \(t_1\). The last subinterval is numbered \(n\), and runs from \(t_{n-1}\) to \(t_n\).

Thus, when the contributions are added in interval order, the finite sum is \(\sum_{i=1}^{n} C_i\). The lower index \(1\) and upper index \(n\) reflect the number and order of the subintervals—not the number of endpoints. The label \(C_i\) is meaningful only together with the interval to which it refers.

Worked Examples

Why Consecutive Changes Add Correctly

When each contribution is the change in a quantity \(Q\) across one subinterval, the contribution on \([t_{i-1},t_i]\) is \(Q(t_i)-Q(t_{i-1})\). Adding these changes makes the intermediate endpoint values cancel. This gives a useful check on whether every interval has been included in the correct order.

Each interior value, such as \(Q(t_1)\), appears once with a positive sign and once with a negative sign, so those terms cancel. Only the final value \(Q(t_n)\) and initial value \(Q(t_0)\) remain. This derivation depends on using consecutive endpoints and including each of the \(n\) subintervals exactly once. It also explains why a sum of signed changes can be checked using the two endpoint values.

A Common Counting Error

Suppose a time interval has the five endpoints \(0,2,5,6,10\). A mistaken count might treat those five endpoints as five intervals and try to assign five contributions. But the intervals are determined by pairs of consecutive endpoints, so they are only \([0,2]\), \([2,5]\), \([5,6]\), and \([6,10]\): four intervals.

For example, if the associated cumulative values are \(8,9,7,10,12\), the four changes are

The correct total has four terms, one for each interval:

A fifth term would have no corresponding consecutive pair of endpoints. The endpoint check confirms the correct result: \(12-8=4\). A reliable habit is to write the endpoint list first, draw or list each consecutive interval, and only then number the contributions.

Keeping the Partition and Sum in Sync

A partition is useful only when its interval list and contribution list stay aligned. Before evaluating a finite accumulation, check these points:

• The first listed endpoint is the start of the full interval, and the last is its end.
• The endpoints are in increasing order along the interval.
• Each subinterval uses two consecutive endpoints.
• A partition with \(n+1\) endpoints has \(n\) subintervals and \(n\) contributions.
• The first contribution belongs to \([t_0,t_1]\), and the last belongs to \([t_{n-1},t_n]\).

This bookkeeping is not merely notation. It specifies which portion of time or distance each quantity describes. Once the partition is clear, later methods can assign or estimate a contribution for each piece. If the partition or its indexing is wrong, even correctly computed individual values may be added to represent the wrong interval.