tdr22 ————— Graphes quantiles-quantiles

tdr22. —————. Graphes quantiles-quantiles. D. Chessel A.B. Dufour & J.R. Lobry. —————. Histogrammes

29-Mar-2022 des tests de normalité (cf tdr22.pdf 'Graphes quantiles-quantiles'). Avant de commencer les simulations bien s'approprier le contenu de la ...

01-Feb-2017 la droite de Henry se référer à la fiche5 « graphes quantiles-quantiles » pour plus de détails (soyons résolument rétrogrades).

Graphics provide arguably the most common way to carry out these univariate assessments While there are many plots that can be used for distributional exploration and assessment a quantile-quantile (Q-Q) plot (Wilk and Gnanadesikan 1968) is one of the most common plots used

De?ning Quantiles The other quantiles of 1 3 2 2 2 7 3 1 3 3 3 7 can be obtained by linear interpolation between the values of the table The median corresponds to a sample fraction of 5 This lies half way between 0 4 and 0 6 The median must thus be 5×2 7+ 5×3 1 = 2 9 The lower quartile corresponds to a sample fraction of 25

Quantiles via Optimization Quantiles seem inseparably linked to the operations of ordering and sorting the sample observations that are usually used to de?ne them So it comes as a mild surprise to observe that we can de?ne the quantiles through a simple alternative expedient as an optimization problem Just as we can de?ne the sample

Other quantiles sometimes mentioned are percentiles often used about scores for an individual on a standardized exam The pth quantile is the same as the 100pth percentile I looked at several statistics books searching for precise de?nitions of sample quantiles Many books have no words beginning with q in their subject indices

The package Qtools provides functions for unconditional and conditional quantiles These include methods for transformation-based quantile regression quantile-based measures of location scale and shape methods for quantiles of discrete variables quantile-based multiple imputation restricted

REGRESSION QUANTILES' A simple minimization problem yielding the ordinary sample quantiles in the location model is shown to generalize naturally to the linear model generating a new class of statistics we term "regression quantiles " The estimator which minimizes the sum of absolute residuals is an important special case