Element order consistency

How an occurrence is read

Occurrence

An occurrence is one kvg:element inside a jōyō character, under the character group. The character’s own outer group is not an occurrence of itself. A host is a jōyō character in which the element occurs. 林 is one host of 木 and two occurrences.

Split parts

A group with no kvg:part is one occurrence. Groups that carry kvg:part and share kvg:element and kvg:number are joined. Nested copies of one part number are one part. A new occurrence starts when the part number stops increasing in stroke order. 歌 has two occurrences of 丁. The sequence uses the strokes that belong to the occurrence, in stroke-number order. A stroke of something else that falls in the gap is not part of the sequence, and it does not make a second occurrence.

Nested occurrences

A group with no kvg:part whose strokes are exactly the strokes of two or more groups of the same element inside it is a container. It is not counted. The groups inside it are the occurrences. 林 is two occurrences of 木, not three. An element inside a different element is counted on its own.

Sequence

The sequence is the kvg:type of those strokes, as the file writes it. A trailing letter stays. ㇑ and ㇑a are different sequences. A slash stays. ㇔/㇏ is not ㇏.

Common sequence

The common sequence is the sequence with the most occurrences of that element. When two sequences occur equally often, the common one is the sequence whose earliest occurrence comes first in the Gallery list, and then the one with the smaller first stroke number. The component consistency page compares an occurrence with the element’s own file. This page does not. The common sequence here is the sequence the occurrences use most often.

Share differing

The share is the number of occurrences that differ, divided by the occurrences of that element. The table shows that share in thousandths, rounded half up. The sort uses the exact ratio. When two ratios are equal, the element with more differing occurrences comes first, then the one with more distinct sequences, then the one in more characters, then the one with more occurrences, then the element’s code points.

Which elements

The share that differs

How many sequences

Elements by the share that differs

Hosts are not added down the column. One character hosts many elements. Distinct sequences are not added either. The share in the last row is the differing occurrences divided by all kept occurrences. It is not the sum of the shares.

Three pairs