Binary functions and their applications (Q1188715)

The binary functions studied in this book are functions of the form \(f: \times_ 1^ n M_ i\to\{0,1\}\), where each \(M_ i\) is a finite subset of \(\mathbb{R}\) (as a matter of fact, the existence of a linear order on \(M_ i\) is the only property of \(\mathbb{R}\) which is actually used). Chapter 1 is a very short introduction which presents the contents of the book. A generalization to binary functions of the classical theory of prime implicants of switching functions is the subject of Chapter 2: implicants, prime implicants, representations of binary functions by implicants (prime implicants), reduced (i.e., irredundant) and minimal representations (i.e., involving the least number of prime implicants). Chapter 3 presents the dual theory of prime implicates. Chapter 4 gives methods for obtaining reduced and minimal representations. Chapter 5 introduces discrete functions, which generalize binary functions to the effect that the codomain \(\{0,1\}\) is replaced by \([0,\infty)\) (here again the linear order of \(\mathbb{R}\) is the only property which matters). Every discrete function has several representations in terms of binary functions. The theory of implicants (prime implicants) is extended in a natural way to discrete functions and as a matter of fact, the implicants (prime implicants) of any discrete function \(f\) are related to those of the binary functions representing \(f\). Special attention is paid to the prime-implicant representation of monotone discrete functions. The dual theory of implicates (prime implicates) is stated explicitly. Chapter 6 sketches the prospect of applying discrete functions to the description of the state of a technical system depending on \(n\) subsystems and to the classification of a set of objects according to several attributes. We reproduce from Chapter 1 the presentation of the contents of Chapters 7 and 8: ``The structure of binary functions suggests to consider --- in Chapter 7 --- more generally a class of finite Boolean algebras generated by \(n\) partitions of the unit element \(\Omega\). Each element of such a finitely generated Boolean algebra has representations by implicants and implicates deduced immediately from the implicants and implicates of a corresponding binary function. Moreover we may define a probability measure on the generated Boolean algebra. Finally in Chapter 8 the results of Chapter 7 will be used to give representations of the sets (events) of a finitely generated set algebra by implicants and implicates. Furthermore we may give such representations for indicator functions, classes of logical propositions and truth functions''. The book includes also a bibliography consisting of 9 books, a list of symbols and a subject index. Reviewer's remarks. 1. The theory of discrete functions and their applications have been extensively studied by \textit{M. Davio}, \textit{J.-P. Deschamps} and \textit{A. Thayse}. Their monograph, \textit{Discrete and switching functions} (1978; Zbl 0385.94020), is quoted in the bibliography of the book under review, but except for a vague sentence on page 56, there is no concrete reference to that monograph in the text. As a matter of fact, the subject treated in Chapters 2-5 of the present book is studied in the monograph quoted above in much more detail and in the general case of discrete functions. The present author fails to specify whether Chapters 2-5 contain results of his own; anyway, the basic definitions and results in these chapters can be easily located in the cited monograph. 2. Chapters 7 and 8 contain many redundancies, to the effect that the author, who avoids introducing the concept of isomorphism, describes separately several isomorphic Boolean algebras. 3. Definition 8.3.1 of a propositional calculus is unsatisfactory. Firstly it is not clear whether or not the last paragraph on page 123, starting with ``We consider propositions as statements\hbox{\dots'',} is included in that definition. If the answer is yes, then the definition is contradictory because it begins with ``A system \({\mathcal R}\) of any things --- called propositions --- \hbox{\dots''.} Otherwise, i.e. if Definition 8.3.1 consists of just one paragraph, then what the author really defines is the much more general universal-algebraic concept of algebra of type (1,2,2,2,2). The definition intended by the author consists probably of Definition 8.3.1 (one paragraph) together with a function \(t: {\mathcal R}\to\{F,T\}\) satisfying the conditions in Definition 8.3.2, while the paragraph mentioned above should be viewed as a comment. 4. Definition 8.5.1 the Theorems 8.5.2-8.5.3 are also fallacious. There are only two possible interpretations of the author's \({\mathcal T}({\mathcal R})\): either as the function \(t: {\mathcal R}\to\{F,T\}\) or as the range of that function. In the former case Theorems 8.5.2 and 8.5.3 have no content, while in the latter case Theorem 8.5.2 reduces to the trivial remark that the two-element Boolean algebra can be made into a propositional calculus in the author's sense (cf. 3 above) and Theorem 8.5.3 becomes a very unappropriate way of saying that the set \(\{F,T\}\) can be made into a Boolean algebra. 5. Sometimes the author's terminology does not meet the usual conventions: indicator instead of characteristic function, subjunction and bijunction instead of implication \(\to\) and equivalence \(\leftrightarrow\), respectively (on page 123). The author recommends that the latter function be read ``if and only if \dots then'' and claims that ``\(\leftrightarrow\) does not mean equivalence'' although Definition 8.3.2 confirms that \(\to\) and \(\leftrightarrow\) are the conventional functions.