I already shared the Entire playlist of Big Data. My teacher also created the college notes from your videos 📸. Because the content of teacher notes and your videos are exactly the same.
Write the script to sort the values contained in the following vector in ascending order and descending order: (23, 45, 10, 34, 89, 20, 67, 99). Demonstrate the output
U class is very neat and clean...i have dought sir....in 5th rule how can u select the bucket length...in some cases 2bits,4bits,and last 8bits of length...how it is possible
Thank you so much for the thorough explanation! what if we wanna use this algorithm for estimation? I've seen they add half of the size of the partial bucket to the result. Can you please explain?
sir,till when we need to add 1 to the end to form buckets...what is the terminating condition of adding 1 at the end. Because if there is no terminating condition then the answer of the query i.e. how many 1's are there in the last 16 bits.......will be different for different students.
For the case of first 20 bits, the answer would be 13! As you know for the first 18 bits there are 9 bits of one. Now if you increase the count to first 20 bits. The length of entire next bucket (4) will be counted. That is why 9+4=13 is the answer... As DGIM algorithm is meant for providing only approximated count of 1's in a stream data as any streaming data is much huge and continuous. Therefore it's difficult to get correct results! Hope it helps❤️ Best wishes for your exams💕
I have a question. Do we have to always show those 3 scenarios : if new entering bit is 0,1,1 and then we have to count the number of recent/first bits whatever is asked? or can I count it directly in the question's given stream?
Its not mandatory to show all these three scenarios and then count the bits.. it's totally up to you, but if time permits in the exam then please show all these scenarios. Hope it helps! Thanks for watching ☺️
If new number is entering then we have to take it into the right side ? And in book the formula of 1’s is half of the size of oldest bucket+size of all newer bucket.
See standard way of adding elements is from right side and also make sure if you are adding elements from right then you have to start creating buckets from right side only In the video I have told how to count 1's you can use that
Thanks a lot for appreciating!❤️ See for counting 1's in a stream we use DGIM algorithm because it gives the approximated answer by forming buckets. For counting 1's in a specific window also you can use DGIM algorithm. But since the window contains finite number of elements you can use any simple algorithms also for counting. In short, both the problems are going to count the 1's but the DGIM algorithm has a sublinear time complexity, which makes it efficient for counting 1's in large windows / stream. Hope it helps!❤️ Best wishes for your exams.... 💕
Heyy that's a great question, look once buckets are formed it's location/timestamp is also stored.. so if you want to count number of 1's in a particular location/timestamp you can easily traverse to that timestamps and access the number of buckets that comes under it and easily count the total number of 1's in it.. Ans: Here, for your example, for getting 1's from first 18 bits, you can start counting from the left and easily count it because here the data is extremely small and it's easy to access. But when it comes to big data, the counter will traverse to the target timestamp and then will count the first 18 bits.... Hope it helps! Thanks for watching! Please do share with your friends ❤️😁
Look in the first 18 bits there are total 9 1's, And now if you ask for first 19 bits you can see that the 19th bit lies in a separate bucket with length 4 So the total number of 1's in the first 19 bits will be 9 (for first 18 bits) + 4 (bucket length that contains the 19th bit) = 13. So in summary the answer will be 13. Now this is acceptable because dgim is meant for counting "approximate" numbers of 1's. Hope it helps, please share it with your friends too ♥️☺️
For this case yes you are correct! As DGIM algorithm is meant for providing only approximated count of 1's in a stream data as any streaming data is much huge and continuous. Therefore it's difficult to get correct results! Hope it helps❤️ Best wishes for your exams💕
