CHPTER2多维数据图示分析(多元统计分析课件-西安交通大学严明义).doc

CHPTER 2 Exploring Multivariate Data Graphically Introduction According to Chambers et al.《Graphical Methods for Data Analysis》,1983, “ there is no statistical tool that is as powerful as a well-chosen graph”. Certainly graphical presentation has a number of advantages over tabular displays of numerical results, not least in creating interest and attracting the attention of the viewer. And graphical analysis often provides the vehicle for discovering the unexpected. During the last two decades a wide variety of new methods for displaying data graphically has been developed; these will hunt for special effects in data, indicate outliers, identify patterns, diagnosis models and generally search for novel and perhaps unexpected phenomena. Large numbers of graphs may be required and computers will generally be needed to draw them for the same reasons they are needed for numerical analysis, namely that they are fast and they are accurate. So, because the machine is doing the work the question is no longer “shall we plot? ” but rather “ what shall we plot?” There are many exciting possibilities including, of course, dynamic graphics ( see Cleveland, W.S. and McGill, M.E. ,1987, 《Dynamic Graphics for Statistics》). But exploration might begin, at least, with some simpler, well-known methods. Univariate marginal views of multivariate data might, for example, be obtained using simple histograms, stem-and-leaf plots or box-and-whisker plots. But variable- by- variable examination of multivariate data cannot reveal any relationships among variables. Consequently, a better place to begin our discussion of graphical methods for such data is the simple scatterplot. The Scatterplot The simple xy scatterplot has been in use since at least the 18th century and has many virtues –indeed, according to Tufte, E.R.(《The Visual Display of Quantitative Information》,1983) The relational graphic –in its barest form the scatterplot and its variants –is the greatest of all graphical desig