Re: Atomic Structures

From: vldm10 <vldm10_at_yahoo.com>
Date: Sat, 12 Dec 2015 14:58:24 -0800 (PST)
Message-ID: <700669ad-486e-49d8-8c0f-2010c46a320a_at_googlegroups.com>

> The atomic structure of one attribute of one entity has the following
> format:
>
> {identifier, attribute} ... (a)
>
> This format (a) was written as a set. This set is the scheme for the
> identifier of an entity and one attribute of that entity.
>
> Note that one entity can have n attributes. If the above-written entity of
> formula (a), have 5 attributes, then I would have presented the entity as
> follows:
>
> {identifier, attribute1},
> {identifier, attribute2},
> {identifier, attribute3}, .... (b)
> {identifier, attribute4},
> {identifier, attribute5}.
>
> Scheme (b) means that we have made 5 special atomic structures that connects
> the same "identifier" of one entity.
>
> Note that existing data models, represent scheme (b) with the following
> format:
> {identifier, attribute1, attribute2, attribute3, attribute4, attribute5}
> that is as one set.
>

-------------Databases that maintain current state---------------- 

Decomposition of an entity with five attributes into atomic structure (b) is provided with "Simple Form", which I presented in May 2006 at http://www.dbdesign10.com .

An entity in my solution is defined. Note that Codd did not define entity in RM / T, although it is the main structure, which was realized as (invisible) surrogate. P. Chen also did not define the entity in its ER model.

---

(Definition of Entity)                    ...              (*)

I have define the entity as follows:
1. By applying the Leibniz's Law
2. I introduced an extension of Leibniz Law. I introduce the intrinsic and

   extrinsic properties of entities.
3. I use Frege's definition of the concept and extensions. 4. I introduce the identification of attributes, objects and relationships

   as well as states. The identification solves fully Russell paradox.    Therefore Frege's definition of the concept becomes OK.

---

 

So I think that the basic things are precisely defined. That's why I can not accept your claim that I am not precise. As far as I know, no one has defined the entity. There is only a definition, which states that an entity is determined by the attributes.

The text above is only for databases that maintain current states.

---

                   General databases that maintain states of the entities
=========================================================================

Unlike databases that maintain only current states, General database maintained current and past states. General database also can solve a number of other things, that database for the current state can not do. Atomic structure of General databases are different. I have introduced identifiers of states of entities and relationships. Now an entity has the following scheme:

{IdentifierOfState, IdentifierOfEntity, attribute1 ,..., attribute5}.

Atomic structure that correspond to above scheme, now becomes:

{IdentifierOfState,  IdentifierOfEntity}
{IdentifierOfState,  attribute1}
{IdentifierOfState,  attribute2}
{IdentifierOfState,  attribute3}                 ...           (c)
{IdentifierOfState,  attribute4}
{IdentifierOfState,  attribute5}

With each identifiers and attribute can be associated the appropriate knowledge. More details regarding the "knowledge" can be found on my website.
(here I used this example with five states of entities)

We do not define the entity, but states of entity. In fact the entity does not exist, there are only states of an entity. Therefore, in the above (*) definition of entity, we need to define the states of the entity. So in my opinion, here is the first time precisely defined the entity.

Note that the atomic structures are not the same as the binary relations. Atomic structures are significantly different from the binary structures.

Note that Leibniz defined the entity (ie the object) by using their properties. Here objects are not defined as in physics or chemistry. They are not defined by the molecules and atoms, electrons and protons, etc.

Note that here Relational Algebra is quite different. Here we have only states of an atomic relations, that is the relation with a single attribute.
Note that Codd has no above-mentioned atomic relations of (c) in RM / T. 6NF also does not work with the atomic structures from (c). But the atomic relations from (c) are the most important. Note that mapping between data models has not been done in the ER model nor in RM / T. My atomic structures, allow mapping between data models. Far as I know, this is the only solution that provides a mapping between the data models.

So mapping between data models and work with the states of entities and relationships are my answer to your question why atomic structures are important. I wrote on this user group also about other important applications of atomic structures.

Vladimir Odrljin Received on Sat Dec 12 2015 - 23:58:24 CET