Read the explanation
The saved explorer generates assumed distributions rather than loading an observed demographic dataset. Year factor is cohort year minus nineteen seventy divided by fifty six. Income boost is the positive part of income ratio minus one, times point four five. The distribution center is minus two point five plus one point two times year factor, plus twice income boost, minus point four times stigma. With illustrative stigma zero and income ratio one, its center moves from minus two point five in nineteen seventy to minus one point three in twenty twenty six. Negative gap means the man is older in this source. Bars compare the absolute magnitudes at one hundred pixels per year: two hundred fifty and one hundred thirty. This visual explains chosen coefficients, not measured historical changes or evidence about individual relationships. The curve samples gaps from minus fifteen through fifteen in half year steps: sixty one bins. It modifies positive gaps with an autonomy factor, and modifies the media curve differently for gaps at least three and gaps below zero. It then normalizes each curve by its own half step weighted sum. Doubling raw mass in one region therefore does not simply double its final normalized probability. In a hypothetical two region example with raw masses one and one, each receives one half. Multiplying only the first region by three yields three quarters and one quarter. Bars compare the first region probabilities at four hundred pixels per probability unit: two hundred and three hundred pixels. This illustration explains normalization, not an actual media statistic. The exported aggregates integrate these sampled bins; the tail thresholds include exactly five and minus five, while peers include absolute gaps up to two. The age pairing matrix uses ages twenty through sixty every two years, producing twenty one values on each axis and four hundred forty one tiles. A distance from the selected age centers determines a Gaussian shaped intensity. Positive gaps multiply by the larger of point zero five and point nine times autonomy minus point five times stigma. Gaps below minus two get a different multiplier; all remaining gaps, including small negative gaps, get one point three. For a hypothetical pre multiplier intensity one and autonomy one with stigma zero, positive gap intensity becomes point nine, while the peer branch becomes one point three. At two hundred pixels per intensity unit those bars span one hundred eighty and two hundred sixty pixels. The color domain is zero to the largest tile, not a sum normalized probability. Brightness is relative within that matrix and is not an observed pairing frequency.