Grouping of data is entailed to facilitate the process of data analysis in the great numbers, especially the data whose has fluctuated condition and multi spectral properties. This study aims to create a concept of grouping data with several new approaches in the analysis of its data grouping. The signal fluctuation used is HF (high fluctuation). There will be a way of approach to represent the pattern into one value that can represent data from the pattern. Furthermore, a representative point for each set of data is set and a unique code is made to compare between one data set in 1 material with data sets contained on different materials. The materials objected to the study are H2O and H2O mixed with NaOH. We analyze the data set by performing the mean values, standard deviation values and variance to mean ratio (VMR) values. Then, their goals are to gain, which dataset groups have the fairly stable value and also can represent the data set for subsequent processing. The results show that the groupings of data sets per 200 data sets are sufficient to represent a good percentage value. In other world, it is noticeable to say that we can obtain the characteristics of a material based on its fluctuation pattern.