I'm new to ER diagrams. I noticed that draw.io (which was recommended on Stackoverflow) does not have a one (optional) to one (mandatory) relationship.
Let's say, I have two tables "user" (id, affiliate_id) and "affiliate" (id). There doesn't have to be an affiliate, in which case user.affiliate_id would be null.
However, if there is an affiliate, then user.affiliate_id will link to affiliate.id.
So wouldn't that be a one (optional) to one (mandatory) relationship?
PS: I was thinking that maybe user.affiliate_id must not be null in a strict sense. However, it doesn't break the foreign key constraints (at least for SQLite 3).
I think you are describing a one-to-many relationship. One user can (optionally) be associated to one affiliate, but the same affiliate can be associated to more than one user.
Or am I misunderstanding?
Yes, you are right, that would be a zero/one (or zero/many) to one relationship type, which has to be shown in the diagram. There are different notations for ER diagrams (and it's therefore, in fact, preferable to use UML class diagrams). For instance, in the notation used by Oracle the optional end of the connection line representing the relationship type is annotated with both a zero and a one symbol. In UML, the annotation of the optional association end would be 0..1 (if single-valued) or 0..* (if many-valued).
Related
I am learning about ER modelling about Database systems. My problem is that there is a entity called books,enitity named user and I want to create a borrows relationship between user and book with attribute issue date. I modelled it as described but it was pointed that borrows cannot be a relationship because a same user can borrow a book twice. Can anybody explain me what this is as I am using issue data as an attribute so records in borrows relationship would not collide as I will use PK as userid,bookid and issue date. How can I model this accurately? I am a little confused in this.
In the ER model, entity relations consist of attributes of a single entity set, in which the PK identifies only one entity set. Relationship relations have a composite PK that represents two or more entity sets.
Your question uses a composite PK that represents two entity sets (userid and bookid) and a value set (issue date). Strictly speaking, it's neither an entity relation nor a relationship relation. It's a combination of a relationship relation (two entity keys) and a weak entity set (issue date functions similar to a weak key). If we want to be creative, we might call it a weak relationship.
If I was forced to draw an ER diagram for this, I might present it like this:
The ER model isn't a complete logical model (unlike the relational model) and there are some situations which aren't handled well or at all. This is one of those situations.
As per description, User and Book are the entities.
One user can borrow an instance of book.
Similary, one user can borrow multiple instances of book, whether It can be same
instance or various instances.
So every transaction between the User and Book has the Issue Date.
Neither the user nor the book has the Issue Date.
Here, the relationship between User and Book are Many to Many.
The Bridge table is Transaction. We can name it as Borrow also as per your interest.
Now, The user has one to many transactions.
Every Book has one to many transactions.
Every transaction is a combination of a User and Book.
Note: Since every user can have the same book multiple times and at the same day. So we can have a composite primary key of user_id, book_id and Issue_timestamp as there is a chance of redundancy in the Issue Date in the same combination.
As i know optionality means the minimum cardinality of a relationship which is denoted as optional to optional, mandatory to optional, mandatory to mandatory..
And Participation denoted as bold line and a normal line.
In the Internet some refer participation as the dependency of the entity to the relationship which is also looks like identifying and non identifying relationship.
and some refer it as the minimum cardinality
What is the correct definitions of those relationships and what is the difference..
Let's start with definitions and examples of each of the concepts:
Total and partial participation:
Total participation (indicated by a double or thick association line) means that all the entities in an entity set must participate in the relationship. Partial participation (indicated by a single thin line) means that there can be entities in the entity set that don't participate in the relationship.
Medicine participates totally in the Produce relationship, meaning that Medicine can't exist unless Produced by a Laboratory. In contrast, a Laboratory can exist without Producing Medicine - Laboratory participates partially in the Produce relationsip.
Mandatory and optional roles:
In a relationship, roles can be optional or mandatory. This affects whether a relationship instance can exist without an entity in a given role. Mandatory roles are indicated with a solid association line, optional roles are indicated with a dotted line.
Roles aren't often talked about in database tutorials, but they're an important concept. Consider a marriage - a relationship with two mandatory roles filled by the same entity set. In most relationships, the entity sets also define the roles, but when an entity set appears multiple times in a single relationship, we distinguish them in different roles.
In the example above, a Patient can Purchase Medicine with or without a Prescription. A Purchase can't exist without a Patient and Medicine, but a Prescription is optional (overall, though it may be required in specific cases).
Identifying relationship / weak entity:
A weak entity is an entity that can't be identified by its own attributes and therefore has another entity's key as part of its own. An identifying relationship is the relationship between a weak entity and its parent entity. Both the identifying relationship and the weak entity are indicated with double borders. Weak entity sets must necessarily participate totally in their identifying relationship.
In this example, a Prescription contains LineItems which are identified by the Prescription's key and a line number. In other words, the LineItems table will have a composite key (Prescription_ID, Line_Number).
For examples of non-identifying relationships, see the previous examples. While Medicine participates totally in the Produce relationship, it has its own identity (e.g. a surrogate key, though I didn't indicate it). Note that surrogate keys always imply regular entities.
Mandatory/optional vs total/partial participation
Mandatory or optional roles indicate whether a certain role (with its associated entity set) is required for the relationship to exist. Total or partial participation indicate whether a certain relationship is required for an entity to exist.
Mandatory partial participation: See above: A Laboratory can exist without producing any medicine, but Medicine can't be Produced without a Laboratory.
Mandatory total participation: See above: Medicine can't exist without being Produced, and a Laboratory can't Produce something unspecified.
Optional partial participation: See above: A Prescription can exist without being Purchased, and a Purchase can exist without a Prescription.
That leaves optional total participation, which I had to think about a bit to find an example:
Some Patients Die of an unknown Cause, but a Cause of death can't exist without a Patient Dying of it.
Total/partial participation vs identifying/non-identifying relationships
As I said before, weak entity sets always participate totally in their identifying relationship. See above: a LineItem must be Contained in a Prescription, it's identity and existence depends on that. Partial participation in an identifying relationship isn't possible.
Total participation doesn't imply an identifying relationship - Medicine can't exist without being Produced by a Laboratory but Medicine is identified by its own attributes.
Partial participation in a non-identifying relationship is very common. For example, Medicine can exist without being Purchased, and Medicine is identified by its own attributes.
Mandatory/optional vs identifying/non-identifying relationships
It's unusual for a relationship to have less than two mandatory roles. Identifying relationships are binary relationships, so the parent and child roles will be mandatory - the Contain relationship between Prescription and LineItem can't exist without both entities.
Optional roles are usually only found on ternary and higher relationships (though see the example of patients dying of causes), and aren't involved in identification. An alternative to an optional role is a relationship on a relationship:
By turning Purchase into an associative entity, we can have it participate in a Fill relationship with Prescription. To maintain the same semantics as above I specified that a Purchase can only Fill one Prescription.
Physical modeling
If we translate from conceptual to physical model (skipping logical modeling / further normalization), making separate tables for each entity and relationship, things look pretty similar, though you have to know how to read the cardinality indicators on the foreign key lines to recover the ER semantics.
However, it's common to denormalize tables with the same primary keys, meaning one-to-many relationships are combined with the entity table on the many side:
A relationship is physically represented as two or more entity keys in a table. In this case, the entity keys - patient_id and cause_of_death_id are both found in the Patient table. Many people think the foreign key line represents the relationship, but this comes from confusing the entity-relationship model with the old network data model.
This is a crucial point - in order to understand different kinds of relationships and constraints on relationships, it's essential to understand what relationships are first. Relationships in ER are associations between keys, not between tables. A relationship can have any number of roles of different entity sets, while foreign key constraints enforce a subset constraint between two columns of one entity set. Now, armed with this knowledge, read my whole answer again. ;)
I hope this helps. Feel free to ask questions.
I'm searching for a solution to model a "can-be" relationship.
E.g.: A User can be Special User.
Any suggestions?
Types for entities can be viewed (results) as one of the following two things:
Relationships
Fields
The types that results in fields are usually seen as functional redundancy, where different rows/records in a table can have the same type. For example, the sex/gender of a user, or access privileges or permission level from a user as well.
If, however, this field has attributes/properties, so we have a new entity. In this case, users would be relating to, for example, address, which may have the street name, house number, among other things. Or, in your case, if Special User has attributes/properties, we have a "Special Users" entity. The cardinality here depends on how many types a user can have, and that takes into account the problem context.
As you say that a user "can be", it means it "can not be" too. You could:
Set a nullable functional redundancy field in Users entity or,
Set an entity "Types of Users" and create a relationship N-N for
"Users". In this case, you would have an intermediate table that
would allow or not the connection between the two entities.
In both approaches you have the ability to add new types of users in the future without effort.
There is a third case, however, referred to as partitioning, where we have types and subtypes of entities. In this case, entities subtypes inherit fields of their super entities. I believe this is not your case here.
If you have any questions, please comment and I will answer.
I need to create an ER diagram and a relational model for a hospital. I have kept it simple to 3 entities. Please can someone have a look and tell me if I have normalised this correctly?
I am not sure it there should be a : relationship between the trust and patients? Does a person assume that the first entity is a single unit, in this case a hospital has x relationships to....
Diagram 2:
ER Diagram & Relational Schema
Your normalization isn't correct. First of all, you don't list functional dependencies. Without that, any attempt to normalize is just guesswork.
Now, I can make a reasonable guess, but even with that you have some problems. The Hires and Cares for relations aren't reflected in your 1NF/2NF/3NF tables. In your 2NF, you introduce a Shift relation that was presumably derived from Doctor in 1NF, but Patient_ID wasn't present in the latter, unless that's actually what the Shift column was for, but renaming fields isn't part of normalization. Also, what happened to Hospital in 2NF? It reappears in 3NF but with a Location attribute that was derived from where? Also, you're not indicating a primary key for Shift, is it correct to assume the combination of all 4 fields constitute a candidate key? It's good practice to be explicit about your keys.
Can an entity be without attributes?
Your Hospital relation does have an attribute - Hospital_ID, which represents an identity relation. Remember an attribute is a binary relation, and Hospital_ID -> Hospital_ID qualifies.
I am not sure it there should be a : relationship between the trust and patients?
What is the trust? It isn't indicated in your diagram at all.
Does a person assume that the first entity is a single unit, in this case a hospital has x relationships to....
Many-to-many relationships are decomposed into two or more one-to-many relationships, and in reading the cardinality indicators we tend to put the singular entity first, but this isn't a rule or a safe assumption. Be explicit about cardinalities.
In SQL you can describe a binary relation with a table like
Husband | Wife
We know that an husband can have only one wife, and viceversa, so that's a 1:1 relationship, and you can specify costraints such that if you add an husband that is already in the table you get an error, right?
If you add a third column like this
Husband | Wife | Country
We know that in some country one husband can have many wives; now you cannot put easy costraints, you have to deal with the third column.
So from a binary relation we get a ternary relation with different behavior that depends on the third column.
This example is stupid and useless, do you know any other example?
(other example of ternary relationship such that one of the column changes the tuple behavior?)
Thank you.
EDIT:
Another point of view to see my problem:
You have any binary relationship, within a domain: do you know any binary relation that changes costraints (or behavior) as domain changes?
Another example might be that you can apply coupons towards an order, but for certain coupon types you can only apply one per order whereas other coupon types may be combined.