LEAST SQUARES ADJUSTMENT OF DISSIMILAR QUANTITIES
J. E. Lilly
Abstract
J. E. Lilly
Abstract
In the least squares adjustment of observations on quantities of different physical dimensions (such as angles and lengths) it is proposed that the concept “sum of the squares of the residuals” be replaced by the concept “sum of the squares of the reduced residuals”, where “reduced residual” is defined as the ratio of the residual to the probable error. This is equivalent to using weighted squares of residuals, and regarding weights as dimenflional quantities instead of pure numbers. The treatment is applied to a few simple cases, with satisfactory results.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In the least squares adjustment of observations on quantities of different physical dimensions (such as angles and lengths) it is proposed that the concept “sum of the squares of the residuals” be replaced by the concept “sum of the squares of the reduced residuals”, where “reduced residual” is defined as the ratio of the residual to the probable error. This is equivalent to using weighted squares of residuals, and regarding weights as dimenflional quantities instead of pure numbers. The treatment is applied to a few simple cases, with satisfactory results.
Key concepts: Residual sum of squares, Residual, Least-squares function approximation, Mathematics, Explained sum of squares, Lack-of-fit sum of squares, Non-linear least squares, Total sum of squares