Log-linear plots for data representation
โ Scribed by James C. Holste; Kenneth R. Hall; Gustavo A. Iglesias-Silva
- Publisher
- American Institute of Chemical Engineers
- Year
- 1996
- Tongue
- English
- Weight
- 128 KB
- Volume
- 42
- Category
- Article
- ISSN
- 0001-1541
No coin nor oath required. For personal study only.
โฆ Synopsis
When developing models to represent experimental data, residual plots are effective for assessing quality of fit, and these present the differences between data and reference data or a model. However, when some ordinate values are excessive, it is impossible to maintain definition for small values. This happens when highly accurate and precise data arc compared to much less accurate and/or precise data. It is customary to present two plots: one for near zero ordinate values and one for high ordinate magnitudes. We present a plotting method which covers the entire ordinate range on a single plot with adequate definition in all regions.
๐ SIMILAR VOLUMES
A pedigree model for binary data, motivated by log-linear modelling, has been developed to examine evidence for familial aggregation in disease status. From an epidemiological point of view a convenient way to express disease concordance between a pair of relatives is in terms of the odds ratio. For
This paper proposes a method for the representation of functions describing a measured geophysical property (via sparsely scattered ordensely defined point data) by tensor and triple product interval B-splines (IBS). The spline representation facilitates archiving, data storage reduction, visualizat
Many current methods of analyzing band recovery data are described by the classic monograph Statistical Inference from Band Recovery Data ยฑ A Handbook of Brownie et al. (1985). A log-linear model with restricted minimum modiยฎed transformed chi-square estimation and Wald type testing are proposed. Th