Plots are among the most important and widely used tools for scientific data analysis and visualization. With a plot (a.k.a. range group-by query) data are divided into a number of groups, and at each group, they are summarized over one or more attributes for a given arbitrary range. Wavelets, on the other hand, allow efficient computation of (individual) exact and approximate aggregations. With the current practice, to generate a plot over a wavelet-transformed dataset, one aggregate query is executed per each plot point; hence, for large plots (containing numerous points) a large number of aggregate queries are submitted to the database. On the contrary, we redefine a plot as a range group-by query and propose a wavelet-based technique that exploits I/O sharing across plot points to evaluate the plot efficiently and progressively. The intuition behind our approach comes from the fact that we can decompose a plot query into two sets of 1) aggregate queries, and 2) reconstruction queries. Subsequently, we exploit and extend our earlier related studies to effectively compute both quires in the wavelet domain. We also show that our technique is not only efficient as an exact algorithm but also very effective as an approximation method where either the query time or the storage space is limited.