If you think Bogo sort is bad, you haven't heard of miracle sort. This works by checking if the array is sorted, if it's not: check again. You have to wait for some miracle to happen, so the memory is getting corrupted in some way.
At university, we once created a similiar algorithm (funny enough without even knowing that miracle sort was a thing), but based on putting our memory next to a radiation source to basically create our own cosmic rays lol
Bogo sort is both the slowest and fastest sorting algorithm. Depending on your luck you can have it sorted at the first try or never, and me personally I like those chances
@@mihaimanole2643 I think it is more like 1/infinite if the random sequences are not stored. You might never find the correct sort after a long time but you should get there with infinite time!
The best algorithm I know is the Stalin sort. It simply removes the elements that are not in order. Not only is it an O(n) algorithm, but it also SAVES memory when you run it! It's really great!
@@RaefetOuafiqo It is O(n), because you only need a single for loop, for each element check if the previous element is less than the current one and then yield item, otherwise you continue with the next one.
You should have covered quantum bogosort: 1. Use some quantum source of randomness (e.g. radioactive decay) to shuffle the list. If we assume that the many-worlds interpretation of quantum mechanics is true this will result in one universe for every possible list order. 2. If the list is not sorted, destroy the universe (this is left as an exercise to the reader). 3. The only surviving universe is one in which the list is sorted. This sort is not stable, but this problem can be easily fixed: just use the same randomness source to generate random source code and destroy the universe if the generated code is not a stable quantum bogosort.
You should have mentioned Quantum Bogo sort, splitting the universe into all possible combinations of the shuffled array and destroying every universe where its not sorted. It has an O(1) time making it the best sorting algorithm
@@deep.space.12 Yupp. Anything it would gain through optimization still requires it to itterate through the list atleast once. As in, the shuffling alone requires you to atleast go through the items in the list once. And big O notation only cares about how much each additional item increases the time needed. As in, an algorithm that takes 2n and 20000n would both become O(n) in duration. Just as a an algorithm that completes in 10 cycles regardless of input vs one that completes in 20000000000 cycles would both be considered O(1).
I made little markdown explanations in Obsidian for most of these algorithms because our professor had a very questionable teaching method that resulted in like 60% of students confused and clueless. Shoutout to the big brains behind mermaidJS, I wouldn't be able to explain merge sort without your graphs!
@@bruvhellnah I just used the top-down graph to illustrate merge sort splitting and forming back together. I'll try to translate the files this weekend!
Sorry guys, I got my hands full. The script is a mess to translate and I just don't have the time but thank you for showing interest. Maybe I'll start writing a new script in English to cover C/C++, JS, and similar stuff and concepts. I guess It will come in handy to someone in the future.
The great thing about quantum sort is that somewhere in a parallel universe, your data array already exists in a sorted form, and a parallel copy of you is already using it. 😁
It's also worth considering how data might be added to the array causing a need to re-sort. If data is rarely changed, the sort algorithm doesn't need to be most efficient. If data is regularly added or removed, how it's added/removed can be important combined with the sorting algorithm.
indeed same with counting sort, although that's for fairly small numbers (counting sort doesn't scale well as the largest number in the array gets large, which is why radix sort exists)
@@dealloc 1) This can be solved if your data can be transformed into integers in a 1-to-1 correspondence. 2) Only important if you care about positioning.
8:55 I actually needed this video for my Calculator, as I needed to sort positions by there Z value, as fast as possible and with the less amount of ram usage
Put every pixel at the first screen row inside an array, then get the size, then use array.at(size/2) to get the center of the screen, then get the div width divide by two, now use position absolute top: array.at(size/2) - divwidth/2. Congratulations you reinvented the wheel!
Radix Sort, doesn't split in buckets of 10, the general meta is to use a base of 256 and there are two implementations, one with buckets and one with counting, I recommend creel's video on the topic, it's really descriptive.
Your Bogosort implementation is even more stoopid than it needs to be 😂 The sorted() function at 8:17 could simply return false immediately, but instead it assigns false - then pointlessly checks the entirety of the rest of the array for no reason! Well done 👏
Three sorting algorithms can teach you about practical, theoretical, and then mind-bending sorts. Whether you're preparing for interviews -- at any level -- or curious about the topic, try these: Timsort, a hybrid of insertion and merge sort, may have executed in production more times than any sorting algorithm ever. Variations of it are the standard library for Python, Java, and Rust. It works, and those two are a gentle introduction to sorting. Bucket sort rips up the math you learned, if any, in Timsort (by executing in O(n) time). It can be helpful for interview questions. You might never implement this, but it's a lovely idea. Bitonic sorting is what happens when you take advantage of parallelism -- and the result is awesome! The speed-up is better than linear (e.g. 2 cores running 2x fast), if that's what you were expecting. Look it up!
The quicksort algorithm implemented in Haskell: qSort :: Ord a => [a] -> [a] qSort [] = [] qSort (x:xs) = qSort small ++ [x] ++ qSort big where small = [a | a
Oh can imagine it now, senior coming up to me yelling at me, why did i implement this sort when its so slow. And i just turn towards him and say: "I am feeling lucky today".
I once wrote a sort where I would check the array for consecutive numbers, break if the next is lower than the previous and put all these blocks in a new array. Then run over the array again inserting values at their "correct" position in the new array.
Quantum Sort is the clear winner, the array is both sorted an not sorted at the same time, so you simply return the sorted array. Order of magnitude is O(-1)
If your sorting positive integers, you can use the integer as an index into a counting array initialized to zeroes, then make a single pass to count how many times each number occurs, then restore the array from the counting array. You make 3 passes. Pass 1: Init counting array to 0's, Pass 2: count how many times each value occurs, Pass 3: Rewrite array from counting array. If you have negative integers, find the largest negative number, then use the positive value to offset all of the numbers. When you restore from the counting array, subtract the offset. Finding the correct offset is an additional pass. The offset can work in the opposite direction if the smallest number is very large. Offset the smallest large number to 0. This sort is good when the range of numbers is small. The larger the range, the larger the counting array needed.
Great Video Sir , I appreciate you for all the knowledge given by you and for all the efforts that you put in these videos. Sir ,I just wanted to mention one thing that as Books refer to some knowledge and for Indians, book refer to our Goddess Sarawati. So just requesting you not to tear them like this.
Bucket sort is a solid O(n) sort that everyone should know about. It has limited use cases. Another sorting technique is to take advantage of multiple cores and sort in parallel. A good default choice for sorting is quick sort as it will often be more than fast enough. If you want to have some fun. Ask any programmer how long it will take to sort one million integer or float values. The answer is almost always a surprise. Hint, it is fast enough that when processing tons of data a sort then process algorithm is often a big performance improvement over other options such as tree based systems.
Did you hear about CNsort? The Chuck Norris Sort Algorithm (CNSort) is a groundbreaking sorting method that operates on the principle of sheer intimidation. In the world of computer science, algorithms are supposed to logically organize data, but CNSort takes a different approach. Here, arrays don't dare to be unsorted. As soon as the CNSort is invoked, the elements in the array glance up to find Chuck Norris staring them down. Overwhelmed by his formidable presence, the elements immediately line up in perfect order, each one too afraid to be out of place. Efficiency is key: CNSort achieves a sorting time of O(1), because no element wants to waste Chuck Norris's time by being out of order.
These algorithms are all way so efficient! I have written an algorithm with worse time complexity than tree(n). Who doesn't want to make more than grahams number of comparison just to sort two elements?
Insertion sort should be O(N log N). I agree that your implementation of it is O(N^2), but it's very easy to make an insertion into a sorted array an O(log N) operation. Edit: to be clear this can't be done with arrays, because while searching a sorted array is still O(log N) insertion is O(N). So you'd have to copy over the data to a balance binary tree and then copy it back, trading space complexity for time complexity.
_"if you're preparing for a technical interview, these visualizations and audiblizations will absolute change your life"_ Interviewer: "Can you briefly describe to me a Quick Sort?" Me: *_wooOOOP wooOOOP_*
My favourite sorting algorithm is `Stalin sort` : Simply iterate over each element in the list. If the current element is smaller than the last, _simply remove it!_ Stalin sort will always produce a sorted array in O(n) time complexity!
JetBrains: Our new Development IDE - Fleet is the future VS Code killer Jeff: I'm gonna destroy the fleet's career JetBrains: *Made him the new business model
Well, Bogo sort is the theoretically fastest algorithm :) It's also the one that in theory could not be done before the heat death of the universe if you are unlucky enough.
You can also create a single-function Quicksort. Here is my implementation in Lua I always use (note that in Lua arrays start in 1 and not in 0): function Quicksort(arr, L, R) local l = L or 1 local r = R or # arr local p = arr[(L+R)//2] if l < r then while arr[l] < p do l = l + 1 end while arr[r] > p do r = r - 1 end arr[l], arr[r] = arr[r], arr[l] end if L < R then Quicksort(arr, L, r) Quicksort(arr, l+1, R) end end
If you think Bogo sort is bad, you haven't heard of miracle sort. This works by checking if the array is sorted, if it's not: check again. You have to wait for some miracle to happen, so the memory is getting corrupted in some way.
Ah, so I follow Miracle Sort to sort out my real life problems.
Its the "are we there yet?" of sorting algorithms.
"is it sorted yet?"
"no?"
"is it sorted now?"
"still no?"
Expecting heat death
Waiting on those cosmic rays lol!
At university, we once created a similiar algorithm (funny enough without even knowing that miracle sort was a thing), but based on putting our memory next to a radiation source to basically create our own cosmic rays lol
Bogo sort is both the slowest and fastest sorting algorithm. Depending on your luck you can have it sorted at the first try or never, and me personally I like those chances
@@mihaimanole2643 I think it is more like 1/infinite if the random sequences are not stored. You might never find the correct sort after a long time but you should get there with infinite time!
50% chance every run. the array is either sorted or it isn't
welcome to the game of life
@@gabriel837 biggest brain logic I ever saw 😂
@@MusicBox.Melodies No, each time you shuffle you have a 1/n! chance to have it be sorted.
The best algorithm I know is the Stalin sort. It simply removes the elements that are not in order. Not only is it an O(n) algorithm, but it also SAVES memory when you run it! It's really great!
Lemme look that up, I was under the impression that Mao sort is the best algorithm ever
it is O(n²)
That would be aladeen sort sir
Doesn't it send them to the Gulag?
@@RaefetOuafiqo It is O(n), because you only need a single for loop, for each element check if the previous element is less than the current one and then yield item, otherwise you continue with the next one.
My man out here being sponsored by jetbrains itself. Bravo man 👏🏻👏🏻👏🏻
You should have covered quantum bogosort:
1. Use some quantum source of randomness (e.g. radioactive decay) to shuffle the list. If we assume that the many-worlds interpretation of quantum mechanics is true this will result in one universe for every possible list order.
2. If the list is not sorted, destroy the universe (this is left as an exercise to the reader).
3. The only surviving universe is one in which the list is sorted.
This sort is not stable, but this problem can be easily fixed: just use the same randomness source to generate random source code and destroy the universe if the generated code is not a stable quantum bogosort.
Algorithm timestamps:
1:37 - Bubble sort
2:33 - Insertion sort
3:32 - Selection sort
4:11 - Merge sort
5:26 - Quick sort
6:53 - Radix sort
7:54 - Bogo sort
I love how you sorted the time stamps for each.
now we need Beyond Fireship to copy paste this into the description
haha insertion
Thanks, there was too much prelude and sponsor time.
clutch
Mistake at 4:08 where you say merge sort results in quadratic time when you're talking about selection sort
I heard this too
@@ashe_neko I HEARD...something...too
Job interviewers should understand the last line of this video.
"So basically, everything you learned in this video is useless on a practical level"
I was thinking the exact same thing. Learn this stuff to find a job and the shelf it until time to look for the next job, lol.
You should have mentioned Quantum Bogo sort, splitting the universe into all possible combinations of the shuffled array and destroying every universe where its not sorted. It has an O(1) time making it the best sorting algorithm
reminds me of Stalin Sort
Another good one is sleep sort. It’s just sleeps every value.
Wouldn't it still require O(n) time to check if the array is sorted?
@@deep.space.12 Yupp. Anything it would gain through optimization still requires it to itterate through the list atleast once. As in, the shuffling alone requires you to atleast go through the items in the list once. And big O notation only cares about how much each additional item increases the time needed. As in, an algorithm that takes 2n and 20000n would both become O(n) in duration. Just as a an algorithm that completes in 10 cycles regardless of input vs one that completes in 20000000000 cycles would both be considered O(1).
@@deep.space.12 nah, cus then we could just do a quantum bogo search and destroy the universes where the sorted list wasn’t selected. Simple.
This is like learning the most boring topics in the form of a kindergarten musical math lesson. Absolutely loved it !
It's funny how my window screen hasn't loaded yet and there's your voice. Thanks for sharing.
I made little markdown explanations in Obsidian for most of these algorithms because our professor had a very questionable teaching method that resulted in like 60% of students confused and clueless.
Shoutout to the big brains behind mermaidJS, I wouldn't be able to explain merge sort without your graphs!
can you share the explanations?
Did you make your own graphs using mermaidJS ?
@@bruvhellnah I just used the top-down graph to illustrate merge sort splitting and forming back together. I'll try to translate the files this weekend!
Sorry guys, I got my hands full. The script is a mess to translate and I just don't have the time but thank you for showing interest. Maybe I'll start writing a new script in English to cover C/C++, JS, and similar stuff and concepts. I guess It will come in handy to someone in the future.
In pandemic i viewed a LOT of sorting videos, they are mesmerizing in their own way.
I like quantumsort. It does nothing but check to see if quantum fluctuations have caused your data to spontaneously sort itself
The great thing about quantum sort is that somewhere in a parallel universe, your data array already exists in a sorted form, and a parallel copy of you is already using it. 😁
Creel's multi part series on sorting algorithms was really excellent and gave a lot of great details, but this was a nice and condense video!
It's also worth considering how data might be added to the array causing a need to re-sort. If data is rarely changed, the sort algorithm doesn't need to be most efficient. If data is regularly added or removed, how it's added/removed can be important combined with the sorting algorithm.
yes, i was thinking same
wow, my favourite programming youtuber shows videos my buddy from university made a decade ago - never expected that 🎉
Thank you for this video, the sound aspect is really interesting .
And the conclusion is gold 😆
Sponsorship!! Congrats Fireship 🎉🎉
your videos are a treasure trove for anyone eager to learn something new!
I think radix sort does extremely well when the input data is massive, like 10M+ values.
indeed
same with counting sort, although that's for fairly small numbers
(counting sort doesn't scale well as the largest number in the array gets large, which is why radix sort exists)
But only if 1) it's an array of integers, and 2) requires intermediate sorts to be stable.
@@dealloc i mean, anything can be converted into ints, though it would mean strange code...
@@dealloc 1) This can be solved if your data can be transformed into integers in a 1-to-1 correspondence. 2) Only important if you care about positioning.
I UNDERSTAND NOTHING
This is timely as I am currently prepping for a tech interview
8:55 I actually needed this video for my Calculator, as I needed to sort positions by there Z value, as fast as possible and with the less amount of ram usage
I work at the library and was looking for better ways to sort books! The intro confirmed I was in the right place!
6:30. Us chilling to watch it on TV. Thanks fireship
You never dissapoint 🔥
Sorting algorithms are satisfying to hear and look at. :)
Because of this kind of videos, I ended up designing my own version of Merge Sort, called YAMS (Yet Another Merge Sort)... It's satisfying.
Interesting, but which array method will help me center this div?
BogoSort
flexSort
gridSort
positionFixed sort :}
Put every pixel at the first screen row inside an array, then get the size, then use array.at(size/2) to get the center of the screen, then get the div width divide by two, now use position absolute top: array.at(size/2) - divwidth/2. Congratulations you reinvented the wheel!
amazing video as usual!!!
by the way at 4:08 you said merge sort but it was the big O notation of selection sort that you were describing
Radix Sort, doesn't split in buckets of 10, the general meta is to use a base of 256 and there are two implementations, one with buckets and one with counting, I recommend creel's video on the topic, it's really descriptive.
this was the best video i have ever seen this hour. nice!
For Quick Sort at 5:59 he forgot to finish the partition() function with “return partitionIndex;”
Great video!
The explanation was on point. you did miss out talking about run time of all sorting methods after selection sort.
Your Bogosort implementation is even more stoopid than it needs to be 😂 The sorted() function at 8:17 could simply return false immediately, but instead it assigns false - then pointlessly checks the entirety of the rest of the array for no reason! Well done 👏
Three sorting algorithms can teach you about practical, theoretical, and then mind-bending sorts. Whether you're preparing for interviews -- at any level -- or curious about the topic, try these:
Timsort, a hybrid of insertion and merge sort, may have executed in production more times than any sorting algorithm ever. Variations of it are the standard library for Python, Java, and Rust. It works, and those two are a gentle introduction to sorting.
Bucket sort rips up the math you learned, if any, in Timsort (by executing in O(n) time). It can be helpful for interview questions. You might never implement this, but it's a lovely idea.
Bitonic sorting is what happens when you take advantage of parallelism -- and the result is awesome! The speed-up is better than linear (e.g. 2 cores running 2x fast), if that's what you were expecting. Look it up!
definitely the best video on sorting ever !
The quicksort algorithm implemented in Haskell:
qSort :: Ord a => [a] -> [a]
qSort [] = []
qSort (x:xs) = qSort small ++ [x] ++ qSort big
where
small = [a | a
Oh can imagine it now, senior coming up to me yelling at me, why did i implement this sort when its so slow. And i just turn towards him and say: "I am feeling lucky today".
I once wrote a sort where I would check the array for consecutive numbers, break if the next is lower than the previous and put all these blocks in a new array. Then run over the array again inserting values at their "correct" position in the new array.
I was always waiting for the day fireship covered algorithms. Did I know I was waiting? No. Was I pleased. Yes indeed
Those animations are trippy. It's like watching and listening to a computer think.
RADIX SORT: A Man, who's work doesn't seem useful, but surprises everyone at the end.
8:55 If schools were always honest while teaching
Good to know the class I'm currently taking is completely useless as well!
Just kidding.
Thanks for this awesome video ahead of my exams next month
Ending was master piece
Nice Video! I wish this Video existed, when i hand Algorithms and Data structures in Unitversity...
i like how he told us at the end all of these r useless
I love this guy. Nothing can prepare you for the end of this video. 😂
With the sound effects for the sorting algorithms, I am reminded of the NES game “Life Force”, which brings back great memories.
Quantum Sort is the clear winner, the array is both sorted an not sorted at the same time, so you simply return the sorted array. Order of magnitude is O(-1)
that ending was glorious
Well, that ESCALATED quickly ! 🤣🤣🤣
this brings me back to when youtube randomly recommended a sound of sorting video
If your sorting positive integers, you can use the integer as an index into a counting array initialized to zeroes, then make a single pass to count how many times each number occurs, then restore the array from the counting array. You make 3 passes. Pass 1: Init counting array to 0's, Pass 2: count how many times each value occurs, Pass 3: Rewrite array from counting array. If you have negative integers, find the largest negative number, then use the positive value to offset all of the numbers. When you restore from the counting array, subtract the offset. Finding the correct offset is an additional pass. The offset can work in the opposite direction if the smallest number is very large. Offset the smallest large number to 0. This sort is good when the range of numbers is small. The larger the range, the larger the counting array needed.
Well that went way over my head.
Bro just got sponsored by Jetbrains 🔥
holy fuck a WHOLE NOTHER FIRESHIP CHANNEL?!
4:07 Minor correction to the viewers. Here he's talking about selection sort and not meme sort. After this, there is the merge sort section
And my dumbass has watched this twice right now and still doesn’t understand a thing but I’m all for it
4:08
It should be "Selection sort also results in quadratic time complexity."
RADIX SORT is like a student, who (seem to) doesn't study all year, but outperforms in Exams :)
Great video!
Great Video Sir , I appreciate you for all the knowledge given by you and for all the efforts that you put in these videos. Sir ,I just wanted to mention one thing that as Books refer to some knowledge and for Indians, book refer to our Goddess Sarawati. So just requesting you not to tear them like this.
2:01 The ES6 way to swap two values in an array in one line with no temp variable: [arr[j], arr[j+1]] = [arr[j+1], arr[j]]
This was awesome, thank you.
This is brutally COOL
0:09 Love how Shellsort is outperforming Quicksort quite dramatically in the example :D
Didn’t understand nothing but the animation are just gorgeous
The end was epic.
Bucket sort is a solid O(n) sort that everyone should know about. It has limited use cases. Another sorting technique is to take advantage of multiple cores and sort in parallel. A good default choice for sorting is quick sort as it will often be more than fast enough.
If you want to have some fun. Ask any programmer how long it will take to sort one million integer or float values. The answer is almost always a surprise. Hint, it is fast enough that when processing tons of data a sort then process algorithm is often a big performance improvement over other options such as tree based systems.
The entertaining world of important refund info
2 years late but appreciating it anyways
Did you hear about CNsort?
The Chuck Norris Sort Algorithm (CNSort) is a groundbreaking sorting method that operates on the principle of sheer intimidation. In the world of computer science, algorithms are supposed to logically organize data, but CNSort takes a different approach. Here, arrays don't dare to be unsorted. As soon as the CNSort is invoked, the elements in the array glance up to find Chuck Norris staring them down. Overwhelmed by his formidable presence, the elements immediately line up in perfect order, each one too afraid to be out of place. Efficiency is key: CNSort achieves a sorting time of O(1), because no element wants to waste Chuck Norris's time by being out of order.
Thanku for posting this video milord
This needs to be updated because of the latest *OpenAi breakthrough* in dorting algorithms 💡
Imagine an alternate universe that's identical to our own, except bogo sort always finishes first try.
Radix sort is my favorite
I loved this video, Jeff plz create a crash course series on youtube on "DATA STRUCTURES & ALGORITHMS" plzZZZ
These algorithms are all way so efficient! I have written an algorithm with worse time complexity than tree(n). Who doesn't want to make more than grahams number of comparison just to sort two elements?
I would also recommend you to explain merge heap in your video.
If you are haiving trouble learning C , then you must create your owm programming langauage
-Fireship
Watching those pancakes getting binned at the end hurt my soul
I hate being that guy but mid should be start-(end-start)/2 to avoid Int overflow. Love the videos, you probably know this anyways.
Insertion sort should be O(N log N). I agree that your implementation of it is O(N^2), but it's very easy to make an insertion into a sorted array an O(log N) operation.
Edit: to be clear this can't be done with arrays, because while searching a sorted array is still O(log N) insertion is O(N). So you'd have to copy over the data to a balance binary tree and then copy it back, trading space complexity for time complexity.
omg I found the second fireship channel
Computer: 'I fear no man, but that thing...'
Bogo sort doing random noise on the screen.
Computer:'... scares me.'
_"if you're preparing for a technical interview, these visualizations and audiblizations will absolute change your life"_
Interviewer: "Can you briefly describe to me a Quick Sort?"
Me: *_wooOOOP wooOOOP_*
My favourite sorting algorithm is `Stalin sort` :
Simply iterate over each element in the list. If the current element is smaller than the last, _simply remove it!_
Stalin sort will always produce a sorted array in O(n) time complexity!
Miracle sort is the goat 🐐
Thank you 🙌
JetBrains: Our new Development IDE - Fleet is the future VS Code killer
Jeff: I'm gonna destroy the fleet's career
JetBrains: *Made him the new business model
0:04 most important scene in this video:
* random
* almost sorted
* reverse sorted
* repeated values
0:10
8:42
You are the best!
AHH!, Finally pure programming (sorting) 🎵 LoFi 🎧 Music (Kind of).
Well, Bogo sort is the theoretically fastest algorithm :) It's also the one that in theory could not be done before the heat death of the universe if you are unlucky enough.
08:21 bogo sort makes some sick beats, tho.
You can also create a single-function Quicksort.
Here is my implementation in Lua I always use (note that in Lua arrays start in 1 and not in 0):
function Quicksort(arr, L, R)
local l = L or 1
local r = R or # arr
local p = arr[(L+R)//2]
if l < r then
while arr[l] < p do
l = l + 1
end
while arr[r] > p do
r = r - 1
end
arr[l], arr[r] = arr[r], arr[l]
end
if L < R then
Quicksort(arr, L, r)
Quicksort(arr, l+1, R)
end
end
I don't even code and I found this interesting
Error at 4:09 "and mergsort also results in quadratic time." I believe you meant to say selection sort rather than mergsort.
Hilarious ending 😂