A function that suggests some rounding degrees of a continuous variable (classically formatted or as interval object). The basic idea is 1. to count which factor is observed in the data more often than expected. 2. to check whether a factor can explain at least three observed piles of observations in the data 3. to check whether a factor explains at least 20 % of observations (additional to previous factors). Factors fulfilling this premises are returned as suggested rounding degrees.
suggest_rounding_degrees(x)A vector or interval object.