Details

Published: Saturday, 24 August 2019 11:46

Hits: 1261

 PHP Hypertext Preprocessor (PHP) : is a programming language that allows web developers to create dynamic content that interacts with databases. PHP is basically used for developing web based software applications. PHP is the most popular scripting language used on servers. The php pgm to run is embedded in html code. When we don't want o have static html pages, we use php to generate dynamic html cod. This php script to generate dynamic webpage is run on server, and then once the html code is generated, it is sent to browser for display.

PHP is open source, and has C like syntax. It was created in 1994, and was originally called Personal Home Page (php). php is most used language for website coding, so it will suffice to learn just this language for building websites with dynamic content. If you want to build static website, just knowing HTML and CSS will suffice, but static websites are not easy to maintain, change, and are not used anymore (except for very small websites).

Official website for php. Here you can find lots of php doc and tutorial: http://php.net

 other good tutorial is from w3schools: https://www.w3schools.com/php/

Installation:

On linux mint, we'll use APT cmd to install php. Latest stable version f php is 7.2.PhP is cripting language, and so can scripts written in php with php interpreter. However, we want php to be able to connect to MySQL database, and as well as communicate with web server pgm to send it the contents to display. So, along with php, we install helper packages.

sudo apt install php libapache2-mod-php php-mysql => This installs php 7.2 + extra modules for it to talk to apache2 and mysql.


It installs these pkg on my system (All pkg installed are 7.2 version, and there are no other versions installed):

no versions: php, libapache2-mod-php, php-mysql, php-common => These pkg w/o version number actually point to 7.2 version

7.2 versions:  php7.2, libapache2-mod-php7.2, php7.2-mysq, php7.2-common, php7.2-cli, php7.2-json, php7.2-apcache, php7.2-readline => these pkg with version number 7.2 are the actual pkg installed.

which php => shows  /usr/bin/php as php binary location. This is a softlink to binary php7.2 stored here.

 php -v => shows php version num as 7.2

There are lots of php modules available. PEAR (PHP Extension and Application Repository is, collection of various useful PHP packages.

Running php:  There are various ways to run php. php is mostly used for server side scripting, where it geberates html o/p. It can also be used as cmd line script, to generate any kind of o/p. We will talk about few common ways to run php:

1. Interactive shell: on linux terminal, we can run php script with php interpreter. Just like any other scripting language, we can open php shell and run php interactively:

php -a => opens php interactive shell. Now we can type any php cmd on php cmd line terminal.

php> echo 2+3; => echo cmds work fine in php

5

php> exit

2. php file: we can have php script in a file, and pass that script to php interpreter, just like with other scripting languages.

test.php => create this file, and put below code in this file (.php extension is not necessary when writing code to run on php cmd line interpreter, it' needed only when running on webserver)

!#/usr/bin/php => 1st line indicates that interpreter used should be php

<?php => Any php code needs to be within "<?php" and "?>" tags. This tell the php engine to execute this code as php code. Anything outside this is treated as regular text, and php engine will just print it on screen as is. So if we omit these tags, php code will not execute at all and will just output the code on screen. It's important to have no space in these tags "<?php" and "?>", (i.enot like "<? php") or else they are not recognized as valid tags. Note: these tags were not needed in interactive shell, as php understands that it'sinteractive session, so every code entered is valid php code.

echo "Hello \n"; => simple echo stmt

phpinfo(); => This is a very useful function provided by php, which provides a lot of info abut php engine that was used to run this script. It is also used to validate if php is installed correctly and is running fine.

?>

any garabge text =>Anything written outside php tags is printed as is on screen. php engine doesn't even look at this and ignores it completely (as it's outside the php tags)

Now, after saving the script, change the permission to executable => chmod 755 test.php

./test.php => prints "Hello". If the 1st line (!#/usr/bin/php) in php file above is omitted, then we need to type "php test.php"

3. webserver script: We can have php file that we can pass to a webserver. Once php is installed on your machine, apache webserver on same m/c is able to access it, w/o doing anything more. The server finds out that this file needs to be interpreted by PHP because we used the ".php" extension, which the server is configured to pass on to PHP. So, .php extension is very important here, as w/o that, the webserver won't know what to do with this file. php files can have regular html code or any text in it. php parser looks for php code inside php tags, and does whatever is asked of it. Parsing in this manner allows PHP to be embedded in all sorts of different documents, as everything outside of a pair of opening and closing tags is ignored by the PHP parser. Then it returns the final generated file (replacing code within php tags with the output of the script) to server engine, which sends it to the browser.

If we have apache installed, we'll see a default file here: /var/www/html/index.html

IF we open browser, and type "127.0.0.1" in the browser address, we'll see default ubuntu page show up (see in apache tutorial section). Now if we put some other file here named test.php, and access it via addr "127.0.0.1/test.php" in the browser, it will be sent to apache server, which will pass it to php engine (since it has .php extension), We can copy below code in test.php file (we'll need to open this file preceeding it with sudo, since only root has permission to edit in this dir).

sudo emacs /var/www/html/test.php

<html>
 <head>
  <title>PHP Test</title>
 </head>
 <body>

  <p>** Outside : before php **</p>
 <?php echo '<p>Hello World</p>'; ?> 

  <p>** Outside : after php **</p>
 </body>
</html>

 Now, on accessing addr "127.0.0.1/test.php", we will see "Hello World" printed on browser, as the php code has been parsed by php engine. We'll also see text with "** ... **"  that's outside php tags.

PHP syntax: We'll be talking about php syntax when used as webserver script. Just like any other language, php file is parsed by parser, which parses tokens and looks for var defn, operators, std keywords, functions, etc. language defines data types as number, string, etc. Variables store different data, belonging to one of these data types. We have operators to operate on these var. We have conditional stmt to get pgm to take diff execution paths.

1. Any PHP file needs to have a .php extension for webserver to recognize it and pass it to php interpreter. php file can have any text in it, but anything within php tags (<?php and ?>) needs to be valid php code, and that is the only code that will be parsed by the php engine. Php syntax refers to php code inside these tags.

2. If php contains only php code, and nothing else, then closing php tags (?>) can be omitted.

3. Each php stmt ends with semicolon ";", (except for the tags. tags imply continuation of stmt, so no ; used. same for stmt like "if else" which are 1 stmt, so no ; used except inside { } which may have multiple stmt in them).

3. comments: C style (// for single line and /* ... */ for multi line) of comments supported. PHP also supports scripting language type single line comment (starting with #).

4. case sensitivity: php variables are case sensitive, but std keywords, class, functions are not case sensitive.For ex, $book and $Book are different, but "when" and "WhEn" are same.

5. whitespaces in php serve as a separator for tokens, so that the interpreter can parse the file for tokens. However, it seems tolerant of non-whitespaces also (i.e ( a < b ) and (a<b) both seem to work ).Not sure, how lenient it is with whitespaces, but it's a good idea to always have whitespaces.

6. variables: variables are specified with $ sign in front of var name. var name can only contain alphanumeric char and a underscore (A-Z, 0-9, a-z, _), and must can't start with a number. variables are case sensitive. We do not specify data type (i.e int, char, etc) of var, php automatically assigns it based on it's value. Thus, it's called loosely typed language, and is similar to almost all other scripting languages, which never assign data types. However, in php 7, data type declaration can be added using "strict" keyword.

php supports following data type:

• string: sequence of char inside single or double quotes. ex: $str = "my name"; MAny builtin func can be used to manipulate string. ex: strlen("Hello world!"); => returns length of string as 12.
  
• integer: $x=435; php func var_dump($x) returns var type and value.
• float: $x=4.23; php func var_dump($x) returns var type and value.
• boolean: 2 values "true orTRUE" and "false or FALSE". $x=true; NOTE: there is no quotes around this (As $x="true" makes $x a string, and not boolean). booleans are usedin conditional stmt
  
• array:stores multiple values in single var.ex: $cars = array("Volvo","BMW","Toyota"); $cars[0] refers to "volvo", so on. can also be assigned as $cars[2]="BM", Many func available to manipulate arrays.ex: count($cars), sort($cars), etc.
  
  • associative arrays: similar to perl associative arrays, used to store key/value pair. ex: $age = array("Jim" => "25", "Joe" => "35"); now $age['Jim'] refers to value "25". This can also be assigned as $cars['Joe']="maruti"; We can loop thru associative arrays using foreach. function print_r($age) explained below, will print array's all key/value pairs w/o going thru "for" loop for each array element.
    
• object: An object is a data type which stores data and information on how to process that data. We use "class" for this, just like in other Object oriented languages.
• NULL: Null is a special data type which can have only one value: NULL. Any variable is created without a value is automatically assigned a value of NULL. ex: $x=NULL;

The scope of var (i.e places where it can be called or referenced) is of 3 types:

• global: var declared outside of any function have global scope, and can be accessed only outside the function (so not truly global, as var not available within the function). However, re declaring these var with "global" var inside the function allows these var to be used within the function.
  • superglobals: there are various predefined superglobal var, that can be accessed anywhere on php code.There are lots of them, and each of these var store useful info in an associative array. ex: $_SERVER superglobal => holds information about headers, paths, and script locations. ex: $_SERVER['HTTP_HOST'] => Returns the Host header from the current request
• local: var declared within a function have local scope, and can only be acccessed within that function.
• static: this is still a local var declared within a func, but we can declare it to be static, which will prevent the var from getting deleted when exiting the func. So, next time func is called, value of the var from last time is still preserved.

7. operators: php divides operators in following groups:

• arithmetic operators: used on numerical values. +, -, *, /, % (modulus), ** (exponentiation)
• assignment opeartors: used to assign values to var. =, +=, -=, *=, /=, %= (a +=b is same as a=a+b)
• comparison operators: ==, === (both equal and same type), !=, !==, <, >, <=, >=
• increment/decrement operator: ++$x (pre increment, increment $x first, then return $x), $x++ (post increment, return $x first, then increment $x), --$x, $x--
• logical operators: and (same as &&), or (Same as ||), xor, ! (not). ex: $a = $x and $y; $b = !$a; $c = $a || $b;
• string operators: . (Concatenation), .=(conctenation assignment, $a .= $b is same as $a = $a . $b). ex: $c = $a . $b (concatenates string $a and $b, to form new string $c, with no space in b/w $a and $b)
• array operators: +, ==, ===, !=, !== these operators same as comparison but apply to array and check key/value pair match. + implies union of 2 arrays
• conditional assignment: ?= (if else) same as in C language

8. conditional stmt:

• if else: ex: if ($cond != 0) { echo "name"; $c=$a; ... }. if-elseif: ex: if (..) { ... } elseif (...) { ... } else { ... } => these don't have semicolon as if else are just 1 big stmt. Inside { }, we should have ; for each stmt. else, elseif are all optional.
• switch: to select one of many blocks of code
• loops: while, do while, for, foreach

9. display: various inbuilt functions avalable to display on scrren. i.e. echo(), print(), print_r().

echo and print are used to display values/text. Both can be used with or w/o parentheses (i.e print "a"; or print("a");). echo works same way as linux echo. ex: echo "My $name"; echo "me" . $name . "and"; print syntax same as echo. One diff is that print returns a value of 1, while echo doesn't return any value. So, print can be used in expressions. Both stmt can contain html markup code in them (i.e print "<b> Hi </b>";). prinrt_r() is used to display values in nice human readable format. print_r can be used to display associative arrays, w/o going thru "for" loop. eg print_r($age); where $age is an associative array.

 10. functions: besides the built in functions, we can define our own function using "function" keyword. ex: function myfunc ($a) { ...return $c; }. Now myfunc can be called as myfunc(5); or $a=myfunc("hello"); Functions may or may not have a return value.

php forms:

• $_REQUEST
• $_POST
• $_GET']

Php cookies and sessions: good link here: http://shiflett.org/articles/the-truth-about-sessions

cookies:

A cookie is often used to identify a user. A cookie is a small file that the server embeds on the user's computer. Cookie is a human readable text file. Each time the same computer requests a page with a browser, it will send the cookie too. The server reads these cookies and determines various info about the user stored in cookies. These cookies need to be stored in client's computer by the server in the very beginning. With PHP, you can both create and retrieve cookie values. It is upon the browser to accept or decline cookies. Some websites demand that you agree to store cookies, or they refuse to show you the webpage.

setcookie(name) => this func sets cookie name. other optional parameters as cookie value, expiry time, etc may be set too, If expiry time not set, then cookie destroyed on closing the browser or that tab of browser.

ex: setcookie($cookie_name, $cookie_value, time() + (86400 * 30), "/"); This has expiration time in future = 86400 * 30 sec = 30 days from current time. (since 1 day = 86400 seconds). To delete a cookie, set it's expiration time in past. setcookie should be the very first thing to do in any php file, before we write out any html code using <html> tag.

superglobal var $_COOKIE stores the value of cookie. i.e $_COOKIE[$name] gives the value of cookie with name $name.

ex: if ( isset($_COOKIE[$cookie_name])) { echo "Value is: " . $_COOKIE[$cookie_name]; } //function isset checks if $_COOKIE[name] is null or has  any value

Note: The value of the cookie is automatically URLencoded when sending the cookie, and automatically decoded when received (to prevent URLencoding, use setrawcookie() instead).

how cookies work:

So, the way cookie's work, is when a browser calls www.xyz.com/index,php, it causes the webserver to process index.php file. index.php has setcookie() function before any <html> tag in the file. This setcookie function is written as some conditional stmt, so that if that request is coming from a client computer for the first time, only then this setcookie func is called. The cookie values are sent by client, in the HTTP header of the request (header has lot of other info also, besides cookie info). $_COOKIE var extracts and stores these cookie values. Depending on whether the header sent by client has "cookie values" in it or not,  $_COOKIE may contain these values or be empty.

ex: if (count ($_COOKIE) == 0) { setcookie ... } else {read cookie values from request sent by browser} => this cmd sets cookie only if number of cookies stored in $_COOKIE var is 0. This code sits before any <html> tag.

This causes the cookie to be set if it was already empty, so that the next response sent by server (which is actually the response to this request sent by client),  will include these cookie name/value pairs in the HTTP header of response. NOTE: $_COOKIE stores cookie values that come in the request header of client. If client header doesn't have these cookie values in it, $_COOKIE will be null. Several copies of index.php may be running on the server at the same time, since many clients may be connecting to the same server, but each of them has their own copy running, and stores the cookie value received from header of client response in superglobal var $_COOKIE.  So, $_COOKIE value is unique for each client connection, even though the same index.php script is running for all clients. If there is some other script as welcome.php running on server, it will have it's own unique values stored in $_COOKIE depending on what the browser sent it. If browser sent same cookie value for other scripts, then $_COOKIE will have same value for other script too.

We can check the request header sent by the client by opening "web developer" in the "open menu" icon on far right of "firefox" web browser. Within "web developer", we will see "storage inspector".  Clicking this will open a new section on bottom of that firefox page. Here we will see a lot of tabs along with "storage" tab. Clicking on storage tab will show couple of sections on right side, one of which is "ccokies". Clicking on "cookies" will show all the cookie values stored by that website on this computer.

If we click on "network" tab instead of "storage" tab, we'll see all the http request/response sent/received by the user computer, time it took, etc. This request/response has multiple lines. First line is the request/response line. 2nd line is required header "Host". After that we can have multiple optional headers. One of such optional headers is "Cookie" for request, and "Set-Cookie" for response. A very good site explaining this is here: http://shiflett.org/articles/the-truth-about-sessions. If we click on the latest http request sent by the browser (shown at the bottom of all requests in the "network" tab of firefox), it will show the http request header sent by the client, as well as the http response header received by the client (which was sent by the webserver after processing the http request sent by the client). In request header, we should see a "Cookie" as one of the headers being sent. It should have name/value pair, if the cookie is set on client's m/c already. If the client m/c is visiting this website for first time, then this info may be absent from http request header. Similarly in the response header sent by webserver, we'll see a "Set-Cookie" header being sent (if setcookie function was called in the php script run by the server, for that request). If setcookie function was not called, then this header will be absent from response. This is a very good way to debug and understand how cookies are being sent and received.

Sessions:

Once the cookie name/value pair is set on user m/c, the cookie is stored on user's computer for ever, until it expires. Sometimes, we do not want the info to be stored for ever. We want info to be stored when the user visits our webserver using a browser, and remain stored, until the user closes the browser or that tab of browser. Also, cookie have another problem, where cookies are stored on user's computer, which makes it a security hole. Cookies do not allow info to be accessed across all pages of a website (or they have to be passed individually to each page, and we have little control on which pages can or cannot access these cookie variables).

These issues are resolved via a session. sessions are similar to cookies in many aspects. session is basically treating website as an application started by a particular user. It stores info related to that user across all pages of a website, so that all pages can uniquely identify the user. It does this by setting a user-key cookie on the user's computer. This cookie name is "PHPSESSID" and it has value that looks something like this: 765487cf34ert8dede5a562e4f3a7e12. It's expiry is set to "session" (i.e expires when user closes the browser or tab). Then, when a session is opened on another page, it scans the computer for this cookie. If there is a match, it accesses that session, if not, it starts a new session.

A session is started with the session_start() function. This causes cookie "PHPSESSID" to be stored with a large random value on user's m/c. This cookie is just like any other cookie, and can be read via $_COOKIE["PHPSESSID"]. IThis cookie is used to uniquely identify the connection of this website to the user's m/c. f this cookie was already set, then session_start() function doesn't set it, but just reads it. Once session_start() function is called, and a session has been started, other session var can now be set. Session variables are set with the PHP global variable: $_SESSION (i.e $_SESSION["user"] = "Matt"); Now we can access this session var "user" in any other script too, by reading $_SESSION["user"]. session_start() function needs to be put before any <html> tags. It's put in all the pages, wherever we need to access these session var. As explained above, session_start() doesn't always start a new session (it starts new session only if it finds no user-key match on reading PHPSESSID cookie),

PHP session expire on closure of window/tab of browser. They also have a default timeout limit of 1440 sec (24 min). So, if session is inactive for more than 24 min, the session var $_SESSION gets unset (i.e it stores nothing). Note that the cookie PHPSESSID still remains valid, and can be read in any php script even after timeout, as long as the browser/tab is not closed.

ex: Below we show 2 scripts. Both scripts can access the same session var.

script_1.php

<?php session_start(); ?> // PHPSESSID cookie is set with random key, if not already set

<html>

<body>

<?php $_SESSION["userName"] = "raj"; // $_SESSION superglobal var is set in this script, but is accessible from any script. This var is set to this value only for this PHPSESSID cookie.

print_r($_SESSION); ?>

</body>

</html>

script_2.php

<?php session_start(); ?> // PHPSESSID cookie is not set here, since the script finds this cookie on user m/c already set. So, it reads the cookie session value, and on a match, starts the same session as in script_1.php

<html>

<body>

<?php

if (isset($_SESSION['$userName'])) { echo "Your session is running " . $_SESSION['$userName']; }

?>

</body>

</html>

sql database:

php allows to connect to and manipulate database. We'll talk about mysql database connections via php scripts. Read mysql tutorial section before reading this section. Before php5, MySQL extension was used to connect to MySQL db, but from php5 onwards, MySQLi (MySQL improved) and PDO (PhP data objects) extensions provided (MySQL was deprecated in 2012). MySQLi can only connect to MySQL db, but PDO can connect to many other db too. However, PDO is more complex, so we'll concentrate on MySQLi only.

MySQLi is object oriented, but supports procedural calls too (i.e can work like a function call too, for folks who don't like to code in object oriented format). We'll focus on using mysqli as a function as that's easier to learn.

The PHP functions for use with MySQL have the following general format −

mysqli_function(value,value,...);

The second part of the function name is specific to the function, usually a word that describes what the function doest The following are two of the common functions:

$connect = mysqli_connect($servername, $username, $passwd); //connects to a db. It has return value of link identifier of mysql db server on success, and FALSE on fail. This link identifier is used to perform any operation on this mysql db server
mysqli_query($connect,"SQL statement"); //performs operations on db

NOTE: we can connect to mysql server only with the username who is listed as user in "user" table of "sql" database. If we haven't added any user to mysql server previously, 
then root is the only user allowed. However, if I try to connect to mysql from php script as "root" user, I get an "access denied error" for user "root@localhost".So, add another user to
 php, and connect as that user. 

ex: The php script below very useful to check if your browser can connect to mysql db on a server. This should be the first script to run on any 3rd party webhost server, on which you have your website and mysql db hosted.

Details

Published: Friday, 23 August 2019 20:11

Hits: 1931

Artificial Intelligence (AI):

branch of computer science with goal of making computers or machines as intelligent as humans. John McCarthy is called the father of AI. He demonstarted 1st AI pgm in 1956 at CMU. AI is dominant in fields of gaming (chess, poker, go), natural language processing (NLP, which understands natural language spoken by humans), vision systems (which can comprehend visual images, video), speech recognition, handwriting recognition, robotics, etc.

AI can be classified in 2 - narrow AI, and general AI. narrow AI refers to use of AI for a narrow purpose as self driving car, smart speaker, etc, while general AI refers to doing anything that a human can do.

AI is basically supervised learning, where after the learning process, a system can be made smart enough where it can produce correct o/p for any given i/p. There is so much interest in AI today, is because traditional AI didn't perform better even when the amount of training data grew. But now with specific branches of AI, we can scale AI performance with the amount of data, i.e more the data, higher the accuracy of prediction. This is specifically true for neural nets, and more so for deep neural nets, which are just one of the approaches of AI.

Machine learning (ML):

AI started as wide field, with thousands of approaches to achieve this intelligence in m/c. All these approaches were being worked until 1980s. However, then a particular subset of AI called machine learning (ML) began to flourish, as it showed much greater promise. It was the practice of using algorithms to parse data, learn from it, and then make a determination or prediction about something in the world. It's algorithmic approaches included decision tree learning, inductive logic programming. clustering, reinforcement learning and Bayesian networks among others. One of the very best application areas for machine learning for many years was "computer vision", though it still required great deal of hand-coding. "computer vision or image detection" was the field of identifying images and videos to make sense of that image. However, it never achieved the goal of general AI, at best it was narrow AI, and even then it was very error prone.

Neural Network (NN) or artificial neural networks (ANN):

A particular algorithmic approach of machine learning, called artificial neural networks (ANN) (or simply NN) had been a topic of research for several decades. They were based on analogy of how human brain works. However, they were mostly shunned by AI research community as never showed much promise of intelligence. The problem was even the most basic neural networks were very computationally intensive, it just wasn’t a practical approach. However, in 2000 as super computers got more powerful, this started to change. ANN were able to learn from a large set of data, and get good results. Previous algo in ML saturated in how well they performed, even after more and more data was used to train them. But ANN scaled with data, and allowed the error rate to keep on going down, as more and more data was used to train them. In 2001, for the first time, human face detection was possible from images. This algo from research came into a product in 2006, when camera started having face detection technology in realtime. This was a major accomplishment. However, with ANN, even these powerful computers weren't able to unleash full power of ANN as they weren't able to handle massive computation. 

Deep learning or Deep Neural network:

Then in mid 2000, the field of ANN got major breakthrough. GPU started getting deployed for doing massive computation needed for ANNs. GPU do a lot of computation in parallel, as compared to CPU which do mostly serial computation. In 2012, Andrew Ng at google (previously a professor at Stanford) took these neural networks, and essentially made them huge, increased the layers and the neurons, and then ran massive amounts of data through the system to train it. In Ng’s case it was images from 10 million YouTube videos. Andrew Ng put the “deep” in deep learning, which describes all the layers in these neural networks. He ran a neural network trained using deep learning algo running on 16,000 CPU cores, that learned to recognize cats after watching only youtube videos, and w/o ever having been told what a "cat" is. For the first time, the promise of "intelligence" was realized. Deep learning is one of the methods applied to ANNs, and can be considered a subset of ANN. It just increases the depth of NN, and by doing that, it makes them deep. This makes NN perform exponentially better. Instead of deep NN, you can design shallow NN too if you have limited computed resource, but they do not perform that well (early ANN were shallow).

Convolution neural network (CNN):

There are various classes of ANNs. The class used in deep learning, that we saw above, is called deep neural network. There are various classes within deep neural network. One that is most commonly applied to analyze images and videos is called CNN. In 2010, ImageNet started in image classification challenge. The error rate in 1st 2 years hovered around 25%. In 2012,  an algorithm dropped error rate from 25% in previous years to 15%. This magical algorithm was CNN, and it started the runaway success of CNN. CNN have since dominated the field, and effort every year has been to improve CNN, rather than spending effort/dollar on alternative algo. We will only focus on CNN from now on. The term "convolution" indicates that the network employs a mathematical operation called convolution. Convolution is a specialized kind of linear operation. Convolutional networks are simply neural networks that use convolution in place of general matrix multiplication in at least one of their layer. Even though CNN is a class within deep learning which itself is a subset of ANN, they have all become synonymous with deep learning, and the 3 terms are used interchangeably (DL/ANN,NN/CNN).

One of the reasons CNN are so popular is that they use relatively little pre-processing compared to other image classification algo. This means that the network learns the filters that in traditional algorithms were hand engineered. This independence from prior knowledge and human effort in feature design is a major advantage.

One very good explanation of CNN is here: https://ujjwalkarn.me/2016/08/11/intuitive-explanation-convnets/

Other link here (read both parts 1 and 2): https://adeshpande3.github.io/adeshpande3.github.io/A-Beginner's-Guide-To-Understanding-Convolutional-Neural-Networks/

3Blue1Brown video series on Youtube are absolutely amazing (4 chapters on NN series): https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi

This site is one recommeneded by 3B1B. It's a completely free online Book with 6 chapters: http://neuralnetworksanddeeplearning.com/index.html

You should look at Andrej Karpathy website: https://karpathy.ai. Lots of fun and life learning stuff. This course by Andrej Karpathy on coding your own GPT (videos build up from basics): https://karpathy.ai/zero-to-hero.html

There are video lectures by Prof Andrew Ng (from Stanford) on www.coursera.org

This is a very good starting point for AI, and I would base my next sections on his lectures on Coursera.

Details

Published: Monday, 12 August 2019 20:38

Hits: 938

Audio/Video Bus Standards:

There are many bus and protcols designed specifivally for transferring audio and video data from one device to the other. Some of the common ones you might have heard are HDMI, MIPI, etc. Mostly, it's the video part that takes most of the bus bandwidth, and so these protocol are also called display protocols. We already looked at Audio only Bus Standards as I2S, TDM, PDM, etc. Before we get into these Audio/Video bus specs, let's get some display panel basics, as that gives us some understanding of how these bits transform to form images on LCD/LED panels that you have on your devices.

Display panel basics:

This link shares some basics:

https://www.nxp.com/wcm_documents/techzones/microcontrollers-techzone/Presentations/graphics.lcd.technologies.pdf

Each memory location on the frame buffer corresponds to a pixel on the LCD panel. A 1024 x 600 resolution display (1280 columns x 600 rows) requires 614400 memory locations, with each location having a number of possible colors. The number of bits needed to describe the available colors is called bits per pixel (bpp). For example, 16 bpp can describe 65536 colors and 24 bits can describe 16777216 colors (known as true color). HDR standard defines 10 bit per color resulting in 30 bits per pixel, or about a billion colors. These are high end display panels. A panel with 614400 24-bit locations requires a 1800 KB frame buffer. The panel is continuously refreshed, typically at around 60 Hz, from the contents of the frame buffer memory. So, this frame buffer will need to be updated with new pixel data at the same rate.

The diagram on link above shows VSYNC signal that needs to be asserted every frame, and HSYNC that needs to be asserted every row. At each pixel clk, we send RGB data.

Resolution is usually described in terms of rows. So, 480p means 480 rows. However, with advent of 4K, we describe resolution in terms of columns. For HD screen (or 720p = 1280 columns x 720 rows), we'll need clcok running atleast 1280x720 times for each frame (assuming RGB bits are driven in parallel, and ignoring porch) => 921.6K => Now the frame needs to refresh @60Hz => 921.6K*60 clk pulses per second = 55.3MHz. If we have only single lane to drive all 24 bits for RGB, then this clk will need to be 24*55.3 = 1.3GHz. This is a very high requirement for 3Gbps for just HD display. Full HD (FHD or 1080p = 1920 columns x 1080 rows, FHD may be called 2K as it has 2K columns. It's 1/4th the resolution of 4K) and 4K (3840 columns x 2160 rows or 4096 columns x 2160 rows) have even higher requirement. In such cases, we use YUV format instead of RGB to reduce BW requirement. We may also use compression, as well as multiple data lanes. Still, clk needs to be in GHz range for 4K. That's why differential signaling used for driving these signals, since at these speeds, data fidelity can't be maintained w/o  differential signaling.

Link that explains more about display interface: https://www.usmicroproducts.com/blog/understanding-display-interfaces

Resolution and Refresh rates supported by display interface:

We assume 24 bits per pixel for a typical display monitor. For a 4K resolution and 30Hz refresh rate with 24 bit color depth, we need to transmit 3840*2160*24*30=6 Gbit/sec. This means the display interface or cable will need to support transmitting data at this high speed. So, we use multiple lanes, usually 4 or more lanes to transfer data. Even that will require clk speed of 6/4=1.5Ghz which is very high speed. We'll look at both low and high speed display interface. For a 8K resolution (7680 × 4320) and 60Hz refresh rate with 24 bit color depth, we need to transmit 7680*4320*24*60=48 Gbit/sec. These require clocks running in excess of 10GHz not only on ASIC, but also on cable or long wires. This is very difficult feat to achieve, but has been done recently as of 2020 to support BW of 100 Gbit/sec.

Display communication protocol:

MIPI (mobile Industry Processor Interface): It's a non-profit org formed by alliance of various big companies. They have spec which define device interface technology. Initially, it was for mobile devices, but now encompasses PC, camera, automotive, etc. They define i/f spec b/w processor and camera, display, RF, battery, etc. Most common i/f:

• Display Serial Interface (DSI):
• Camera Serial Interface (CSI):
• System Power Management Interface (SPMI):

DSI:

DSI specifies high speed differential signaling for display controllers. Bus includes 1 CLK lane and 1 or more data lanes. Each lane is 2 wires (pos and neg due to differential signaling, both for clk and data). All lanes travel from the DSI host to the DSI device, except for the first data lane (lane 0), which is capable of a bus turnaround (BTA) operation that allows it to reverse transmission direction. When more than one lane is used, they are used in parallel to transmit data (i.e 4 lanes means 4 bit data transmitted in parallel).

So for basic 1 Lane DSI, we have 4 signals: clk (CLK_N, CLK_P) and data (DATA0_N, DATA0_P).

The link operates in either low power (LP) mode or high speed (HS) mode. In low power mode, the high speed clock is disabled and signal clocking information is embedded in the data. In this mode, the data rate is insufficient to drive a display, but is usable for sending configuration information and commands. High speed mode enables the high speed clock (at frequencies from tens of megahertz to over one gigahertz) that acts as the bit clock for the data lanes. Clock speeds vary by the requirements of the display.

The communication protocol describes two modes of transmission - video mode and command mode.

1. video mode: this is used to drive frames directly containing each pixel RGB values. This is used when external display doesn't have frame buffer to store frames, so frames are passed directly to display with appr panel timing (panel timing being generated by the processor or SOC). New frames are driven for each panel refresh (i.e 30 or 60 times per second), or else the image will be lost. Image data on the bus is interleaved with signals for horizontal and vertical blanking interval. Image data is only sent in HS mode. When in HS mode, commands are transmitted during the vertical blanking interval.

2. command mode:  This is used to send cmds to update frame buffer in external display. Here pixel data is not driven directly, as frame buffer stores the pixel data and refreshes it. Here panel timing is generated by the display controller  itself, instead of by the processor. New refresh data is send via set of cmds. So this mode is power friendly, as frames don't refresh that often, saving a lot of traffic on i/f. Even when they refresh often, not all pixel data changes, so only few pixels may be transmitted.

In cmd mode, there are 2 sets of instructions - DCA and MCS. The Display Command Set (DCS) is a set of common commands for controlling the display device, and their format is specified by the DSI standard. The Manufacturer Command Set (MCS) is a device-specific command space whose definition is up to the device manufacturer. The packet format of both sets is specified by the DSI standard. Packets are composed of a DataID, Word count, ECC, Payload and CRC. Commands that require reading data back from the device trigger a BTA event, which allows the device to reply with the requested data. A device cannot initiate a transfer; it can only reply to host requests.

CSI:

CSI specifies high speed differential signaling for camera modules. It interfaces processor and camera sensor. Same signals as in DSI.

High Speed Display communication protocol:

Besides the display interface we talked above, we also have high speed display interface that are used in TV, computers, gaming consoles to connect display src to display monitor. Here's a link to Video connectors: https://en.wikipedia.org/wiki/List_of_video_connectors

These transfer data at rate of 10's of Gbit/sec. Some of the common ones are:

1. DVI (digital Visual Interface):
2. VGA (Video Graphics array):
3. HDMI (High Definition Multimedia Interface): Very popular and commonly seen on TV, PC, etc.
4. DP (Display Port): More common on high end gaming PC and gaming consoles. Competes with HDMI.

We'll look at each of these in separate sections.

Details

Published: Sunday, 28 July 2019 20:05

Hits: 1071

wordpress:

wordpress is one of the most advanced blogging software, from which you can build almost any kind of website. Most important, it's open source, so you can see and learn how to code. It's written entirely in PHP language.

porting wordpress:

All of the wordpress files/scripts are contained within a wordpress directory, except for the database. So, when you want to migrate wordpress website from one host to another (for ex when moving from godaddy host to ipage host), it's really simple as you just need to move the whole directory and you're almost done. You still need to move the database separately, as it doesn't reside in the wordpress dir. Once you have moved the database, you will need to modify wp-config.php file in the top level wordpress dir to point to the new database. This is needed as the database location has changed. Change these values in wp-config.php file:

1. define('DB_NAME', 'new_db_name'); => If you kept new database name the same, there's nothing to modify.

2. define('DB_USER', 'new_db_user'); => If you kept new database user name the same, there's nothing to modify.

3. define('DB_PASSWORD', 'new_password'); => enter your new password here.

4. define('DB_HOST', 'name.dbxx.x.abc.com'); => Critical that you modify the sql hostname to match the new host name. You can get the new database hostname from the web hosting provider. Another way to find this

At this point, we are done, migrating wordpress. A very good tutorial on this is HERE .

wordpress structure:

In order to understand basic structure of wordpress, I've listed main dir and files:

1. WEB-INF dir:

2. wp-admin dir:

3. wp-content dir:

4. wp-includes dir:

5. Numerous php files:

index.php is the first file that is accessed in the wordpress dir whenever a visitor visits the website. If you recall, index.php or index.html are the default files that the browser looks for when the url is typed (www.abc.com), with no files specified in the url (www.abc.com is equivalent to www.abc.com/index.html or www.abc.com/index.php).

Theme:

I have used default "Twenty Ten" theme. It's named Twenty Ten, since it was released in year 2010. Every year a new theme is released by Wordpress team named after that year. Currently for year 2020, we have "Twenty Twenty" as the theme.

The very first thing you might want to do is to customize the appearance of this theme. This can be done by going to administrator panel, and choosing Appearance tab on left side. Then click on Themes.

Your current active theme shows up. Click "Customize" on it. A new screen shows up. Click on "Header Image", then "Add new image" and upload your new image. Don't worry about the dimensions of the image, as the menu will allow you to crop the size of this new image to fit on your website. Once Published, the new image will show up on the top bar of your website.

You can further customize what gets placed on what side of page via "widgets". Instead of clicking on "Themes", click on "Widgets" on Appearance tab. You will see lot of widgets for images, archives, calender, etc. You can choose your desired widegts and drop them on right side to wherever you want them to be placed. I placed my pictures and search box, this way, on right side of page.

One good way to learn about themes is to create one yourself from scratch:
good tutorial here: http://webdesignerwall.com/tutorials/building-custom-wordpress-theme