Dominant use pairs

The mean uses per compound in Table V is 1.7. The mean uses per compound can be counted tor each ot the uses, it was shown (9) that mean uses per compound is distributed statistically and the outliers on each side could be identified. The ones with low mean uses per compound were highly specific uses such as antimicrobial, antineoplastics, antihistaminics, estrogenics, etc. The ones with high mean uses were the ones which were present in dominant use pairs or clusters such as diuretics, vasodilator, CNS depressant, etc., or which had old imprecise use descriptors such as sudorific, diaphoretic, dermatoses, etc. 

Use Correlations. A fourth cut through the medical use-frequency hyperspace is the one which involves the derivation of dominant use pairs. Table III lists these use pairs as a function of uses/compound. Table III also lists the frequency of individual use pairs. It was noted earlier that the use pairs in Table III have been grouped by closeness of relation and that in a taxonomic sense, the grouped parts constitute a cluster. 

---