 This is our python code to see if one is in the other is point inside rectangle a left a top b all the other directions if it's true then it returns it otherwise it doesn't so basically it now knows if something is intersecting or not and then it's defining is the point inside the rectangle so it wants to intersect even when the smallest point is inside so then you set up pygame pygame that in it and yeah that's the way you start pygame in any program even if it's imported so my window is 400 megapixels high and 400 megapixels wide that's basically easy Cartesian coordinates so I make these direction variables this is all this all is pygame you have to know this they are just selected to make it look easier these four one is the direction that is left and down which would be what do you say 5 o'clock I guess no no 7 o'clock yeah so download don't write is 3 you have to make sure these are defined otherwise you'll have a hard time just memorizing all these stupid numbers which are so useless whatever okay the speed of moving I've kept it as 4 not too fast not too slow most most people will know these color combinations so in order for me not to type again random numbers I define black is blah blah blah green is this and what is that so basically again you just need to draw the the bouncer and make it move and then make the food just randomly pop up which is easy if you have a good knowledge of Cartesian coordinates so that's it that's that's how simple pygame can get and how useful it can be as you can tell from the program I'm also I'm also a beginner even with Cartesian coordinates yeah so this has been a great opportunity for me to give a little introduction to what I'm doing with Python and how did that you know what I'm doing and how it has been useful for me thank you thank you so much Shashank that was wonderful now we have Praveen Patil he's going to talk about portable science labs and the product name is xpxp's good afternoon after these wonderful sessions by these wonderful kids I'm really very tense now just try to finish everything in four or five minutes this is something about a wonderful kid that with which kids can do wonders like the one which you have just experienced now unfortunately this is the situation in our science education system and we just think of what will happen to a vehicle which has these kind of wheels one that big wheel represents theory and the other one represents experiments it doesn't go anywhere most of the times sciences start from the textbook without giving importance to experiments and real understanding and as a result students fail to correlate their classroom experiences to the real life problems that they encounter in real life and to correct this science learning through exploration and experimenting can be a good leisure but if you want to do that providing affordable low-cost apparatus equipment to the students is a difficult thing say for example if you want to do some simple experiments very simple experiments like generating electrical signal and trying to plot it on a screen so things we need we need a CRO oscilloscope that is a costly equipment students can't afford to have them at home right okay you need different sensors what I'll try to do is I'll use this very simple kit and we'll try to fetch data from the hardware using a very simple Python code and and we'll try to see what kind of signal we are generating so what I basically have here is a simple DC motor maybe in the end I'll show you after disconnecting these things we have a very simple DC motor then a small made in a magnet which will get in any toy shop and a screwdriver so using this I'll create a simple pendulum like thing and then okay the thing is very simple okay when we power this motor using a DC battery that we'll start rotating and without connecting any power so if I just rotate the wheel power should be generated at the back end right so I'm doing that I'm just magnet to the wheel and I'll use this as a pendulum just a second please okay there is a small problem there okay maybe I'll try something else first this wire is disconnected okay I'll just explain what exactly happened maybe a little later I can show it to you because we do not have that time so if I just move this right it's a pendulum and the AC waveform will be generated and that hardware can fetch the data and with a simple Python code we can plot it okay very simple experiment that can be done with different instruments for studying sound waves a concept called interference of sound is very difficult to demonstrate in a normal laboratory so what we need is two sources of sound and to have the input for those sources of sound we need frequency generators function generators and then to plot that we need a CRO but that can be done very easily with this kit okay what I am doing here is I have used two simple buzzers okay they will generate the sound and for activating the buzzers and generating square waves here using the same kit and I'm using a simple mic here to collect the signal and the output of the mic is actually connected to a CRO there is a Python code for CRO and that will plot as a graph I'll just try so I just collect one buzzer to one square generated and I'll say start it generates a sound I hope it is audible right and then if I bring it very close to the mic you can see that wave right okay you can change the frequency here and you can see what exactly is the difference that is happening so for kids it's really wonderful to learn these things by self-expression operator stress the other buzzer also if it works yeah so this is one more buzzer with slight difference in frequency and using 3400 find rate and 3600 okay now what I'll do is okay what you saw here is continuous wave with same almost constant amplitude if I start both the buzzers together then you'll be able to see something else right this is what happens when two ways are having simultaneously medium and they super pose when they go in phase you get larger amplitude when they go out of phase amplitude will be less so like this using this very simple instrument many science physics electronic experiments can be done at home so it's like having your own science lab in your pocket and the code running behind this is from Python and because of Python the things have become very easy for a person like me who has never undergone any training in computer science never learned any programming but within 2-3 days I could learn these simple things and you people are from computer science background you know something about programming you can do others with this in fact our dream was to provide a very much low-cost affordable science laboratory to every student in the world and now one student has helped us to develop Android app for this so with a mobile phone and this it has really become a science lab in the pocket thank you very much okay one second for details you can you can log into xpiles.in that is our main site and project is developed by IUAC inter-university accelerator center of New Delhi thank you very much thank you Paveen next we have Shravanti she's going to talk about email classification my name is Shravanti today for morning we were in open space talking about this we had a lot of talk about this I don't know something to demo from a lot of presentation were amazing but this is just a gist of Gmail API and LDA topic model I don't know how many of you know but two months back in Google I or Google earliest Gmail API so before this all of us were using iMaps to access our email server so Gmail API is very handy it is faster to access the mailbox so that is one advantage and one more is I'm talking about LDA topic modeling so we have a lot of machine learning algorithms so LDA is something which we can use to classify the documents so our problem is basically we are from travel industry most of the travel industry the typical problem is they get a lot of requests through emails so and for the operation team it is always difficult to classify the mails they don't know when is the booking request which one is if somebody is following up or somebody is confirming the voucher or invoice mail so these are different categories we wanted to classify from our Gmail so basically we wanted to group our mails so why LDA because there are quite a number of classification algorithms KNN, support vector machine many but for us we don't have time to label the data and all of the data we had was text so LDA was fitting well for our case and coming to Gmail API I would ask everyone to use this because it is really a cool thing they've done from past two months I don't know how many of you know it it is very handy to use just use your IDs usually it will be me when it is when I'm going through my inbox and you just have to give user ID and using OAP you just authenticate your app and you can write code using any language Python Java PHP anything so for us we were packing all the thread IDs a mail when the when a request came so when a booking request is coming and when it is turning into a booking and whether it is dropped after giving us after we give them some quotes so such things we were able to track with this these are the key resources you can extract from Gmail API you can extract the message text you can take the labels under messages red messages and all that and also you can track the history so if you want to track only the latest mails from past one month or two months something like that and for LDA we have a Python implementation of LDA that is called Janssen and for processing the text we are using NLTK this is my crisp talk I don't have something to demo thank you thank you Shravanthi next we have Ramesh he's going to talk out Ansible okay hi I'm Ramesh I will be talking on Ansible I don't have any slides so I will be just going through the Ansible website Ansible how many of you know what how many of you know puppet okay so Ansible is a simple IT automation tool which is built using Python it uses it doesn't need any client side software to execute execute it it's playbook on the host it simply uses SSH to connect to the host or it uses up on SSH or you can use Pythons Peronical library to connect to the remote host so so using Ansible you can do configuration management you can do deployment you can run add up commands you can also do testing using simply you can do any automation using Ansible so can you okay at the core of Ansible there is something called modules Ansible already comes comes preloaded with 200 plus modules which are written in Python core Python but apart of apart of Python you can also use any language like C Java Ruby even bus the only condition it should not is it should emit JSON output unless unless until you are given just an output Ansible can accept any damn language you whichever you want to use so apart of module so what is at the heart of the at the heart of the Ansible so you you include the modules in the playbook playbook is a simple YAML file which contains tasks and and this task in task will execute your modules so simple task file will look like this so this is a task file you are saying what all has to want to execute on so this is a group name which we will call later so these are the simple tasks you give a human readable name for the task you specify the module in this case it's a YAML and you are saying it to install HDD package and its version should be latest that's all just just these two lines of code will actually install HDD on your remote host using SSH so so for playbooks to execute you need to tell them what are host to execute on so Ansible relies upon an initial file where you pass all your host you can also pass your host groups for on that so you can directly pass the group name and it will run on all your host it's again a simple YAML file which Windows uses a part of it Ansible also provides some cool stuff called plugins plugins it's called the plugins so whenever you execute a playbook if you want to run a separate piece of code like when a when a playbook executes it finishes it's it completes its execution you want to run some you want to notify a group of people so you can write an email a search script which will email the output of the Ansible Ansible execution and it will it will send it send a mail to the users you want to notify or you can add something like logging and other cool stuff so it simply goes like you you execute Ansible task you you pass your playbook and playbook will have your custom modules which will which will run those modules on on the host which you which is specified in your host inventory and it will and it will just and it will give you the output and depending on that you can run your callback callback plugins which will again notify your users or do other cool stuff yeah that's it so one one good thing about Ansible is it compared to other other configuration management tool like puppet self and site it's like it doesn't need any agent right on the client's client side agent it simply uses your open SSH or Peramiko and you can also even have a pool base pool base site Ansible will pull all your playbooks from your Git repository directly so you don't need to SSH it you can set up a current job and it will again it will pull you and playbooks directly thank you we have one more slot available is there anybody else who wants to give a lightning talk right now cool so next we have Shantanu who is going to talk about AWS how many of you have used Amazon web services I'm just going to tell you something about Boto first of all some of the people who don't know about Amazon web services this is how the screen look like when you log in and this is how your S3 look like S3 is a is a service where we store all our all our files Redshift is for big data and easy to for all my servers this is T1 Macro cheapest one and this is M3 to extra large which is very costly if you want to see the build you can also see the way Boto now you can use the same interface to manage your web services all this S3 and Redshift that Boto is much more important see when you say import Boto that model gets imported and you can start working with your Amazon Amazon properties here what I'm doing is that I'm just giving access key and secret key so that I can connect to my account then I am creating three accounts ready redshift connection easy to connection in S3 connection so I'm creating three connections using this Boto model and these are the four these are the methods of Boto so Boto dot connect redshift will connect me to a redshift then I will use this connection object later on to create redshift clusters I will use EC3 EC2 connection model connection object later on to create my servers and S3 connection object for creating uploading S3 files and downloading files and deleting the files just see how it is done this is my I will just first let me show you the EC2 things list all these relations so EC2 connection that I just created take that connection object get all the instances will list all the instances that is all the servers that I am already having that is this list the list of three servers so what I am saying is that instead of going to my user interface through programmatically I am managing my Amazon resources first I import Boto I give access key and secret key to connect to my account I create an EC2 connection object and then I say get all the instances to get all the instances then I will get print reservation so I will just print it so just three lines and I get all the instances which I have I can also get the get only instances and I can also connect them I can also create an instance create an image I can create images start an instance is run instances is the method and uses EC2 connection object and run instances amid is this amid the image that I want to base my EC2 instance so I am creating an instance just using four lines of code I can create n number of servers I can configure the server as well whether I want it 6GB, 4GB, 8GB, 62GB Amazon EC2 has n number of options I can also manage my hub disks I can manage attach volumes I can terminate an instance I can copy a file to S3 and I can delete a file from the S3 bucket and Redshift how many of you are using Redshift? quite a few but Redshift is really a good option for database if your MySQL is not scaling and you have a big data really big data use Redshift and here how you can do it programmatically check the active clusters start a new cluster all here Redshift can be managed like this just four lines of code and you can do anything in Amazon properties using BOTO BOTO is extremely powerful use it you will like it thanks