A review of qcc (2.0.1)


Best of the Control Charting Packages

qcc provides the ability to create all of the common quality control charts, sometimes called process behavior charts, and produces clean, attractive output. The charts covered include X-bar and R, Individuals and moving Range (ImR or XmR), P, NP, U, C, CUSUM and EWMA time-weighted charts, capability analysis charts, Pareto charts and operating characteristics (OC) curves. The package provides various options for controlling how limits are calculated and which of the various "Shewhart's rules" to apply. Output is attractive, with with plots and lines plotted in black on a white data background (or scale-line rectangle). "Out of control" points are plotted in red or orange, depending on the rule that is violated.

The documentation is generally adequate, but not complete, and some of the required input data is non-obvious. In particular, the input data needed for creating a moving range (mR) chart is not explained in the documentation. Additionally, the range and moving range charts are not automatically generated with the X-bar and Individuals chart, respectively. While this arguably provides some flexibility in presentation, the two range charts provide valuable diagnostics that should always be plotted.

Given the broad applicability of process behavior charts to statistical process control and to generally evaluating data homogeneity, this package's feature set and general ease of use make it an indispensable part of my library.

The 2.0.1 package has been verified against data provided by Donald J. Wheeler for the X-Bar and R and ImR charts, demonstrating that the package's default settings correctly calculate control limits using the more robust data range method rather than other, less robust methods.