--- title: "5. Linear regression for discrete distributional data" output: rmarkdown::html_vignette: math_method: mathjax vignette: > %\VignetteIndexEntry{5. Linear regression for discrete distributional data} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8} --- ```{r setup, include = FALSE} knitr::opts_chunk$set(collapse = TRUE, comment = "#>", fig.width = 7, fig.height = 4.5) has_cub <- requireNamespace("CUB", quietly = TRUE) knitr::opts_chunk$set(eval = has_cub) ``` ```{r, eval = !has_cub, echo = FALSE, results = "asis"} cat("**This vignette needs the 'CUB' package (for the `univer` data), which is not installed: code is shown but not run.**") ``` Which aspects of a service drive the global satisfaction of a group of customers? With distributional data the question concerns whole distributions: we want to predict the distribution of the global satisfaction of each group from the distributions of the specific items. ```{r} library(DDAWASS) library(ggplot2) data(univer, package = "CUB") univer$Faculty <- factor(sprintf("F%02d", univer$faculty)) univer$Gender <- factor(univer$gender, levels = 0:1, labels = c("M", "W")) items <- c("informat", "willingn", "officeho", "compete", "global") M <- data2MatD(univer, unit = c("Faculty", "Gender"), vars = items, support = 1:7, min.n = 10) xv <- c("informat", "willingn", "officeho", "compete") ``` ## The two-component model Let $Q_{y_i}$ be the quantile function of the response for group $i$, and $\mu_{x_{ij}}$ and $Q^c_{x_{ij}} = Q_{x_{ij}} - \mu_{x_{ij}}$ the mean and the centred quantile function of the explanatory variable $j$. The model of Irpino and Verde (2015) is $$Q_{y_i}(t) = \beta_0 + \sum_j \beta_j \mu_{x_{ij}} + \sum_j \gamma_j Q^c_{x_{ij}}(t) + e_i(t), \qquad \gamma_j \ge 0 .$$ The $\beta$ coefficients explain the **position** (the means) of the response distributions; the $\gamma$ coefficients explain their **variability and shape**. The constraint $\gamma_j \ge 0$ guarantees that the predicted quantile functions are non-decreasing, i.e. that predictions are proper distributions. Parameters minimize the sum of squared Wasserstein distances between observed and predicted distributions; since centred quantile functions are orthogonal to constants, the problem splits into an ordinary least squares regression on the means and a non-negative least squares problem on the centred parts. ## Estimation ```{r} pars <- WD.regression.two.components(M, Yvar = "global", Xvars = xv) round(pars, 3) ``` The `AV_` coefficients are those of the regression of the mean global satisfaction on the means of the items; `compete` (competence of the staff) has the largest effect, followed by `willingn` and `officeho`. They coincide with an ordinary regression on the means: ```{r} means <- as.data.frame(get.MatD.stats(M, "mean")$mat) round(coef(lm(global ~ informat + willingn + officeho + compete, data = means)), 3) ``` The `CEN_` coefficients show how the variability of the judgements on each item is transferred to the global judgement: again `compete` has the largest weight, while the dispersion of the evaluations of the office hours has almost no role. ## Predictions and goodness of fit ```{r} pred <- WD.regression.two.components.predict(M[, xv], pars) gof <- WD.regression.GOF(M[, "global"], pred) gof$RMSE_W gof$OMEGA gof$PSEUDOR2$index ``` `RMSE_W` is the root mean squared Wasserstein error (in the units of the scale); `OMEGA` and the pseudo-$R^2$ measure the share of the variability of the observed distributions reproduced by the model (Irpino and Verde, 2015). Observed and predicted distributions can be compared for some groups. Predicted distributions are discrete, but their support values are not restricted to the integers 1-7: they are linear combinations of the quantiles of the explanatory distributions. ```{r, fig.height = 5} sel <- c("F02_M", "F06_W", "F11_M", "F13_W") obs <- MatD2long(M[sel, "global"]); obs$type <- "observed" prd <- MatD2long(pred[sel, ]); prd$type <- "predicted" both <- rbind(obs, prd) ggplot(both, aes(x, cdf, colour = type)) + geom_step(direction = "hv", linewidth = 0.8) + facet_wrap(~ row) + scale_colour_manual(values = c(observed = "grey20", predicted = "#c62320")) + labs(x = "global satisfaction", y = "cumulative distribution", colour = NULL) + theme_minimal() ``` A summary view compares observed and predicted medians and interquartile ranges of all the groups: ```{r} q_obs <- sapply(c(0.25, 0.5, 0.75), function(p) get.MatD.stats(M[, "global"], "quantile", p)$mat) q_prd <- sapply(c(0.25, 0.5, 0.75), function(p) get.MatD.stats(pred, "quantile", p)$mat) df <- data.frame(group = get.MatD.rownames(M), obs = q_obs[, 2], prd = q_prd[, 2], obs_lo = q_obs[, 1], obs_hi = q_obs[, 3]) ggplot(df, aes(prd, obs)) + geom_abline(colour = "grey60") + geom_errorbar(aes(ymin = obs_lo, ymax = obs_hi), width = 0, colour = "grey60") + geom_point(size = 2) + labs(x = "predicted median", y = "observed median (bars: observed IQR)") + theme_minimal() ``` ## Bootstrap confidence intervals Standard errors can be obtained by resampling the groups: ```{r} set.seed(42) B <- 200 boot <- t(replicate(B, { idx <- sample(get.MatD.nrows(M), replace = TRUE) WD.regression.two.components(M[idx, ], Yvar = "global", Xvars = xv) })) ci <- t(apply(boot, 2, quantile, probs = c(0.025, 0.975))) round(cbind(estimate = pars, ci), 3) ``` With only 26 groups the intervals are wide. The competence of the staff (`compete`) has a clear effect on both the position and the variability of the global satisfaction; the information received (`informat`) and the willingness of the staff (`willingn`) matter mostly for the variability (`CEN_` coefficients), while the office hours (`officeho`) have a small effect on the means only. ## References - Irpino A., Verde R. (2015). Linear regression for numeric symbolic variables: a least squares approach based on Wasserstein distance. *Advances in Data Analysis and Classification*, 9(1), 81-106. doi:10.1007/s11634-015-0197-7 - Crocetta C., Irpino A., Antonucci L., Marin C. (2025). A performance indicator and its decomposition according to the impacts of different aspects based on distributional data. *Quality & Quantity*, 59(3), 2091-2110. doi:10.1007/s11135-024-01956-0