Physical vs Logical vs Abstract Data Types (DST) and data structure (DS)
All computer science is about abstraction as everything is binary, and that's why its sometimes hard to learn, let's break down this terminology.
Abstract data types
Also known as ADT, let's start our discussion about the primitives or scalar types.
Abstract data types are abstracting away 0 and 1 from us,
Representation and operations, each data type, is represented in some specific way and can have a set of operations to it.
16 bit integer is 15 bits of 1 and 0, and the leftmost bit is whether its a signed or unsigned integer.
Arithmetic Operations can be done on integers addition, subtraction, multiplication division modulos
- Relational operations, less than greater than, increment than decrement..etc
Primitives or scalar types, is the programming language provided types, like integers, strings, floats..ect
We dont concern our selves too much how internally this operations are performed.
When we learn programming languages, we mostly learn about representation of data, mostly scalar types, or primitive types and the operations we can do to them.
So the concept of representation and definition of data, and operations on the. while hiding the internal details is in essence abstract data types
For example in programming languages, we define classes, which has objects, and we give those special behaviours and different details, this is how we define them and there operations. for example we can create a class of linkedlists, with different operations
Add(element, index) or inserrt(element, index)
This is how we create our own abstract data types, but the thing is, there is some very known and efficient data types, we call them data structure, they allow us to efficiently utilize the resources, the main memory in the most ideal way, and we extend the programming languages to perform what we want us, us developers.
Physical data structure
Data structure is collection of data structure, so multiple data stored together.
This data structure defines how memory is allocated, they define how the memory will be organized in order for them to store the data.
Most programming languages supports arrays
its a collection of memory allocation, side by side, back to back.
Arrays are fixed size, or arrays size is static, we cannot increase or decrease an array size after its creation, we can only assign to it, or desctruct it (deallocate the memory)
Arrays can be created in the stack or the heap, we use a pointer to allocate arrays in the heap.
We use arrays when we know the maximum number of elements we need
Lined lists are dynamic by nature, you can add more elements or delete them.
its a collceiton of data, where each piece is called a node
Each node holds data, and a pointer to the next node.
Lined lists live in the heap, the head pointer can be in the stack.
Logical data structure
This data structure, are an abstraction over the physical datastructure, Physical data structure store the data in memory, physically, but the logical data strucutre is the protocol, the discipline or the set of rules and policies we use to use the physical data structures in way we want for better organiztion of our data.
All operations we do to this data structure are defined also by the logic we set. the API.
This data structure can be split into two types, linear data structure, and non linear data structure.
Linear data sturcture
linear data structure include
Non linear data structure
- Hash tables
Big O notatian
Its how we analyse algorithms and operations on them, how efficient they are specially when we approach infinity. in respect of N input.
To measure the performance of our data structure, our job is to find out how much work we do per N elements, means for example if to find a word in a dictionary takes you iterating through every single word, means thats Order of N, if you can find the word by just looking at a fraction of the book or a subset, there is high chances, its a log to the two of N, so time complexity is always in respect of our input, or how much element exists in the data we are working with, the job is not to find the exact time it takes, because that varies from machine to machine or from buzy processors to another..ect Our job is to find the least amount of time we need to process N elements, we analyse the behaviour, how much more time we need to get the task done. a behaviour not an exact time.
Many times, when we do something, we need extra memory than what we are working with, so in that case, we would have to consider that, because we dont always have extra memory, so thats the memory complexity it can be order of N or less or more, depends on the data type or algorithm we are working with.