Visual examination of ceteris-paribus (CP) profiles is insightful. However, in case of a model with a large number of explanatory variables, the people may end up with a large number of plots that may be overwhelming. In this chapter, the authors describe a measure that can be used for such a purpose and that is directly linked to CP profiles. It is worth noting that the larger influence of an explanatory variable on prediction for a particular instance, the larger the fluctuations of the corresponding CP profile. By using the average of oscillations, it is possible to select the most important variables for an instance prediction. This method can easily be extended to two or more variables. In this chapter, the authors present analysis of CP-profile oscillations as implemented in the DALEX package for R. At this point, the authors are not aware about any Python libraries that would implement the methods presented in the chapter.