100+ Computer Science Concepts Explained

what's the first thing you should dowhen your code throws an error obviouslyyou should change nothing and try to runit again a few times if that doesn'twork you're gonna need a computerscience degree the awesome thing aboutsoftware engineering is that you canlearn to code and get a high paying jobwhile literally having no idea howanything actually works it all justfeels like magic like a pilot driving agiant metal tube in the sky whileknowing nothing about aerodynamics[Music]welcome to computer science 101 intoday's video you'll learn the sciencebehind the garbage code you've beenwriting by learning 101 differentcomputer science terms and concepts thisis a computer it's just a piece of tapethat holds ones and zeros along with adevice that can read and write to itit's called a turing machine and intheory it can compute anything like thegraphics in this video or the algorithmthat recommended that you watch it atthe core of modern computers we have thecentral processing unit if we crack itopen we find a piece of silicon thatcontains billions of tiny transistorswhich are like microscopic on offswitches the value at one of theseswitches is called a bit and is thesmallest piece of information a computercan use however one bit by itself is notvery useful so they come in a package ofeight called a byte one byte canrepresent 256 different values like allthe characters that you type on yourkeyboard in fact when you type into yourkeyboard the character produced isactually mapped to a binary value in acharacter encoding like ascii or utf-8binary is just a system for countinglike the base 10 system you normally usewhen counting on your fingers but itonly has two characters one and zerohumans have a hard time reading binaryso most often it's represented in ahexadecimal base 16 format where tennumbers and six letters can represent afour bit group called a nibble as adeveloper when you write code in aprogramming language it will eventuallybe converted into machine code which isa binary format that can be decoded andexecuted by the cpu what it doesn't dothough is store data for yourapplications for that computers haverandom access memory or ram it's like aneighborhood and inside of every houselives a byte every location has a memoryaddress which the cpu can read and writeto you can think of the cpu and ram asthe brain of the computer but in orderfor a computer to be useful it needs tohandle input and output an input devicemight be the keyboard and mouse while anoutput device might be your monitorluckily most developers don't need toworry about how this hardware fitstogether because we have operatingsystem kernels like linux mac andwindows that control all hardwareresources via device drivers now tostart hacking on the operating systemyour first entry point is the shellwhich is a program that exposes theoperating system to the end user it'scalled a shell because it wraps thekernel it takes a line of text as inputand produces an output this is called acommand line interface not only can itconnect to your own computer but withthe secure shell protocol it can alsoconnect to remote computers over anetwork now that you have access to themainframe it's time to pick aprogramming language which is a toolthat uses the abstraction principle tomake computers practical to work withfor humans by simplifying differentsystems layer by layer some languageslike python are interpreted that meansthere's a program called an interpreterthat will execute each line of code oneby one other languages like c plus arecompiled they use a compiler to convertthe entire program into machine code inadvance before the cpu attempts toexecute it this results in an executablefile that can be run by the operatingsystem without any extra dependenciesnow every programming language has avariety of built-in data types torepresent the data we're working with inour code instead of bytes we work withmore human-friendly things likecharacters and numbers now the mostfundamental way to use data in yourapplication is to declare a variablethis attaches a name to a data pointallowing you to reuse it somewhere elsein your code python is a dynamicallytyped language which means we don't needto tell the program exactly which datatype is assigned to a variable it justfigures it out automatically howeverother languages like c are staticallytyped and that means you need to specifythe data type of a variable in your codewhen you define a variable its value isstored somewhere in memory on thehardware and you may need to allocateand free up memory throughout theprogram a pointer is a variable whosevalue is the memory address of anothervariable which can be used for low-levelmemory control many languages don't wantto deal with low-level memory managementand instead implement a garbagecollector which automatically allocatesand de-allocates memory when an objectis no longer referenced in the program[Music]now the data types available aredifferent in every programming languagebut typically you'll find int torepresent whole numbers which may or maynot be signed or unsigned to representnegative numbers as well when numbersrequire a decimal point they typicallyuse the floating point type it's calleda float because there's only enoughmemory to represent a certain range ofnumbers at a certain precision and isbasically a form of scientific notationto make computers faster if you needmore range or precision many languagesalso have a double that doubles theamount of memory used for the number nowwhen it comes to characters you'lltypically find the char data type torepresent a single character or morecommonly a string to represent multiplecharacters together ultimately thesecharacters get stored in a memoryaddress somewhere but they need to bestored in a certain order when the orderstarts with the most significant byteand the smallest memory address it'scalled big endian or vice versa if theleast significant byte is stored in thesmallest address it's called littleendian when it comes to practicalsoftware engineering one of the mostfundamental things we do is organizedata into data structures the mostuseful data structure is probably thearray or list just like a shopping listit organizes multiple data points inorder however it also maintains an indexof integers that starts at zero and goesup for every new item in the list thatcan be useful but you don't actuallyneed an index to create a list of itemsanother option is a linked list whereeach item has a pointer to the next itemin front of it another option is a stackthat follows the last in first outprinciple it's like stacking a set ofplates then when you want to access thedata you pop the last one off the topthe inverse option is a queue which isfirst in first out just like when youget into the red line the first personthere is the first one to be fed nowanother extremely useful data structureis the hash which might also be called amap or dictionary it's like an array butinstead of an index of integers youdefine the keys that point to eachindividual item giving you a collectionof key value pairs in many cases thoughit's not efficient to organize data in alinear way to address that problem wehave trees which organize nodes togetherin a hierarchy that can often betraversed more quickly this cansometimes be too rigid of a datastructure though so instead a graph canbe created to connect multiple nodestogether in a virtually unlimited numberof ways a graph has a node for the dataand an edge for the relationship betweenthe data points data structures areessential but they don't do anything bythemselves to do something useful you'llneed to code up an algorithm which isjust code that solves a problem i tookthe initiative increating the internet in our code wehave several mechanisms for implementingalgorithms the most fundamental of whichis a function which is a block of codethat takes an input then does somethingand returns an output like a variable afunction has a name and it can be calledfrom other parts of your code withdifferent input parameters calledarguments one thing you might do in thefunction body is compare one value toanother every language has a variety ofbuilt-in operators like equality greaterthan and less than that you can use tocompare two values if a is greater thanb then it forms a value of true but if bis greater than a then the value isfalse true false is what's known as aboolean data type and whenever your codeproduces a value like this it's known asan expression but not all code willproduce a value sometimes your code willsimply do something which is known as astatement a good example is the ifstatement which handles conditionallogic for example if the condition istrue it will execute this code otherwiseit will short circuit and run the codeinside of the else block another verycommon type of statement is a loop awhile loop will run this block of codeover and over again until the conditionin the parentheses becomes false thatcan be useful but more often than notyou'll want to loop over an iterabledata type like an array most languageshave a for loop that can run some codefor every object in the array oriterable data structure now in somecases a function may not have an outputwhich is generally called a voidfunction an interesting thing aboutfunctions is that they can callthemselves when a function calls itselfit's called recursion because when donelike this by default it will recurseforever creating an infinite loop thathappens because when you call a functionthe programming language will put itinto memory on what's known as the callstack which is a short-term chunk ofmemory for executing your code when afunction keeps calling itself thelanguage will keep pushing frames ontothe call stack until you get a stackoverflow error to avoid this youralgorithm needs a base condition so itknows when to terminate the loop nowwhen you write an algorithm you'll needto determine if it's any good and thesystem for doing that is called big-onotation it's a standard format forapproximating the performance of analgorithm at scale it may reference timecomplexity which is how fast youralgorithm will run and space complexitywhich deals with how much memory isrequired to run it developers have manydifferent algorithm types at theirdisposal the most crude option is bruteforce where you might loop over everypossible combination to hack somebody'scredit card pin a more sophisticatedapproach might be divide and conquerlike binary search where you cut theproblem in half multiple times until youfind what you're looking for anotheroption is dynamic programming algorithmswhere a problem is broken down intomultiple smaller sub-problems and theresult of each computation is stored forlater use using a technique calledmemoization that means if a function hasalready been called it will use theexisting value instead of recomputing itagain from scratch then we have greedyalgorithms that will make the choicethat is most beneficial in the shortterm without considering the problem asa whole one example of this isdijkstra's shortest path algorithm onthe flip side we have backtrackingalgorithms which take a more incrementalapproach by looking at all the possibleoptions like a rat and a maze exploringall the different potential paths nowwhen it comes to implementing your codethere are always multiple ways to getthe job done one programming paradigm isdeclarative where your code describeswhat the program does and the outcomebut doesn't care about things likecontrol flow this style of programmingis often associated with functionallanguages like haskell the otherparadigm is imperative programming whereyour code uses statements like if andwhile providing explicit instructionsabout how to produce an outcome it'sassociated with procedural languageslike c today most general purposelanguages like python javascript kotlinswift and so on are multi-paradigm whichmeans they support all these options atthe same time in addition toobject-oriented programming the ideabehind oop is that you use classes towrite a blueprint for the data orobjects in your code a class canencapsulate variables which are commonlycalled properties as well as functionswhich are usually called methods in thiscontext it's a common way to organizeand reuse code because classes can sharebehaviors between each other throughinheritance where a subclass can extendand override the behaviors of the parentclass and it opens the door to all kindsof other ideas called design patternsnow a class by itself doesn't actuallydo anything instead it's used toinstantiate objects which are actualchunks of data that live in yourcomputer's memory often you'll want toreference the same object over and overagain in your code when data islong-lived it can't go in the call stackinstead most languages have a separatearea of memory called the heap whichunlike the call stack can grow andshrink based on how your application isused it also allows you to pass objectsby reference which means you can use thesame object in multiple variableswithout increasing the memory footprintbecause it always points to the samechunk of memory in the heap now what'sinteresting is that if we go back to thecpu that we talked about in thebeginning you'll notice that it containsmultiple threads a thread takes thephysical cpu core and breaks it intovirtual cores that allow it to run codesimultaneously there are someprogramming languages that supportparallelism where you can write codethat literally executes on two differentthreads at the same time however manylanguages out there are only singlethreaded but that doesn't mean theycan't do two things at the same timeinstead they implement concurrencymodels like an event loop or co-routinesthat can pause or delay the normalexecution of code to handle multiplejobs on a single thread at the same timenow in modern computing we're rarelyworking with the bare metal cpu and raminstead we work in the cloud with avirtual machine which is just a piece ofsoftware that simulates hardware thatallows us to take really big computersand split them up into a bunch ofsmaller virtual computers these machinesare the backbone of the internet and areconnected via the internet protocol eachmachine has a unique ip address toidentify it on the network that ipaddress is usually alias to a url thatis registered in a global databasecalled the domain name service now toestablish a connection the two computerswill perform a tcp handshake which willallow them to exchange messages calledpackets on top of that there's usually asecurity layer like ssl to encrypt anddecrypt the messages over the networknow the two computers can securely sharedata with the hypertext transferprotocol the client may request a webpage then the server will respond withsome html modern servers provide astandardized way for a client to requestdata which is called an applicationprogramming interface or api the mostcommon architecture is rest where urlsare mapped to different data entitiesavailable on the server and that bringsus to our final topic mother effinprinters you're gonna need to learn howthese things work inside and out becauseevery time you go to grandma's houseshe's going to ask you to fix it whichshouldn't be a problem for a computerscientist like you thanks for watchingand i will see you in the next one