Algorithms: Bit Manipulation

[@griffintoal4410]

still a little 'bit' lost ;)))

[@SumoCumLoudly]

this is not a good explanation video

[@jasperzanjani]

@Juan2003gtr your GTR needs a new transmission Juan, ve te a la chingada

[@pwn0x80]

Yep better watch xorpd assembly lecture

[@ohdude6643]

Boomm drop da maik

[@3deeguy]

@@pwn0x80 Thanks. I just subbed xorpd. I tried to read some code that had a lot of bit shifting.

[@dorondavid4698]

Great video, but the masking section was done too quickly... You needed a bit more in depth examples for each type.

[@FWTteam]

It's all the bits and pieces into one video, represented very interestingly. Thank you!

[@RagingWhoremoans]

I'm still trying to compute all of this information. It's a bit overwhelming.

[@SK-yb7bx]

wah wah whaaaaa

[@ivanleon6164]

shift to another things and then come back to continue w the video.

[@shanmukh111]

It's a "bit" too much I agree

[@SkeleCrafteronYT]

You need the basics of binary to understand. It will help to learn binary conversion to hexa-decimal and base-10. Also learning "signed" binary, will help with understanding the twos complement which I think she explained quite well. You'll get there!

[@enesabdulla5872]

@@SkeleCrafteronYT ​ @Ivan Leon I think the person was making a joke (a 'bit' overwhelming).

[@simplelife8722]

I came to youtube understand bit manipulation after reading first chapter of your book.... And boom I have the author explaining me herself... Thank you soo much Gayle😀

[@KentoNishi]

I'm a *bit* interested

[@esjihn]

just a little bit

[@toddchaney2454]

@@esjihn Not a tad bit though.

[@devprakash5320]

naughtyy!!

[@uncaged3076]

nice

[@laurenortiz2120]

This REALLY helped clarify the shifting and mask business for me. Thanks!

[@HackerUC]

I got a little lost at 7:23 when she starts talking about shifting the number to the left by i spots

[@shanthureddy4234]

yeah exactly she didnt bother to explain why original number % (random) mask should not be zero in the first place ,,,

[@ohdude6643]

What you need to keep in mind is: if you shift too much to the left, you end up with communism. Too much to the right you've got a Facist state. Neither are good.

[@gulammohiddin5747]

The part you mentioned was regarding getting the bit at the 'i' th position. To get that bit we need to do a Bitwise AND with the binary number which has 1 at the 'i'th position. So to create that binary number we take 1, which is represented as 00000001, shifting the 1 to 'i'th position, (which gives 00100000 in the video, actually depends on i) and do a bitwise AND with the binary representation of the given number. If the result is a zero then you know that the 'i'th bit is zero else its a 1.

[@ognimoddd]

don't you mean a little BIT lost?

[@edgar6705]

@@gulammohiddin5747 Hi Gulam, so you always start with one and then start shifting?

[@shankar7435]

After watching several videos of others, I finally landed on the right one, which is precisely what I was looking for. Thanks a lot.

[@quirkyquester]

the best video talks about bit operation i've ever found! Thank you so much!

[@rydmerlin]

It has to be said Gayle is a superwoman. Looking at what she was wearing it appears as though she did this series of videos in the same day and that’s just amazing to me that she can offer this instruction like that.

[@mj-sv7ep]

At 2:45, she says "So let's think about this. What number would we have to add to 00 100 10 to get 1 000 000 0." I know about 2s complement, but I donot get why we need to get 1 000 000 0?

[@led07zeppelin]

1. convert 123 to binary. 2 Invert bits, and then add 1. Why do we add one? 4:22

[@edgar6705]

I think 1(0000001) is the number you need to add to your inverted number so you can get the negative version of 123

[@MonjurulIslamKhan]

For 2's complement, you don't have to do that much work. Just keep same all the bits till you encounter first 1 from right to left. After first 1 complement each bit. For example, 010100 -> xxx100 ->101100. :D

[@Pignemo]

but that's a slower algorithm than inverting the bits and adding one.

[@ghostriley22]

I think it's faster on paper but slower for the computer

[@RickeyBowers]

Subtracting one and then inverting the bits also works - each of these methods have utility depending on the environment. For example, they could simplify a more complex bit twiddling, or have dedicated hardware. Thanks for sharing this alternative.

[@JackWootton]

If anyone tries this in Python, be careful because shifting 1 "i times" will not provide the correct bit mask for the "i th" bit: 1

[@solisoma1012]

let your counting start from 0 down to the left

[@zongweifan2353]

you helped me get much more clear with the bit manipulation things, thanks!

[@cbade993]

She understands English, Spanish, Russian and Korean given the books on her desk?

[@snlgrg]

explanation of concept is very good but only one thing is missing i.e some real world problem, just introduce some real world problem without solution, You can post solution in some other video.. ;)

[@jorgeherrera1074]

go to their website and sign up and you'll have examples, really worth it, just signed up a few days ago, probably the best free help ive found

[@ryantay87]

Completely agree. Been doing the 30 day code challenge to better myself.

[@MarioRugeles]

snlgrg your concern is fair, altough I would also encourage you to dive more into the conceptual approach just for fun. Real world applications will always be there waiting 😀

[@shanthureddy4234]

check out the geeks for geek

[@SoyJayP]

Excellent video. It feels good to finally understand something about Bit Manipulation

[@DarkSpydaIV]

Very wordy and much technical jargon. Hard to follow for someone new to the topic. The curse of knowledge means that the more you know, the easier it is to talk like an expert and alienate non-experts. The way to explain this bit manipulation information to beginners is to keep it in very simple wording. Such is understanding your audience and who is receiving the information. If that audience is people who are new to the topic, they will be lost in translation.

[@emilyhuang2759]

8:19 for big picture of getting, setting, and clearing

[@kevinrodriguez7403]

does she show us how to actually add "1" to the binary inverse of the positive binary representation? like its a pretty big step she just glosses over

[@ashutoshbichare]

Number:23 Left shift:Number*2=23*2=46 Right Shift:Number/2=23/2=11

[@SJHunter86]

Hey :) Just so you can see why its -12: 1 1 1 1 0 1 0 0 in artihmetic base 10 numbers works out to be: -128 + 64 + 32 + 16 + 0 + 4 + 0 + 0 = -12 When the signed bit is a 1, the value is -128, then you add left to right as usual. Hope this helps!

[@OhhOmni]

@8:12 when Gayle is explaining how to "clear the ith bit", she mentions that all you have to do is invert your existing binary number, then AND it with the inverted number. But if you pay attention, you will see that she doesn't invert the two 1's near the right side of the binary number. Does anyone have any idea why she may have done it this way?

[@design48art]

OhhOmni I had the same question. Hopefully someone can chime in with some insight

[@graham5774]

"AND it with a mask with a 1 everywhere else but that one spot". She then said to get this by inverting the previous mask, not the existing number.

[@cooldeep7492]

Jump to the 5:20 if you want to know the difference between logical and arithmetic right shift.

[@kardashevr]

9:00 The only thing I understood 100%

[@ohdude6643]

So that's about what? 2 maybe 5% of the whole?

[@acymiranda]

hahahaha

[@SharronDenice]

Lol this tickled me so

[@satyamsingh4653]

in clearing the bit, the invert bits are wrong at the 3& 4th place?

[@sakshamjain6900]

you guys provide a great explanation, but I still think you need to go a little slower so the people can understand it much better and please try to explain videos in depth a little bit ;)

[@blank336]

so how can you tell the difference of -123 and 133 in binary? they both are 10000101

[@APaleDot]

This is where the distinction between signed and unsigned integers comes in. If the data-type we're talking about is an unsigned 8-bit integer, then 10000101 is 133. If it's a signed 8-bit integer, then 10000101 is -123. Unsigned 8-bit integers can be in the range 0 to 255. Signed 8-bit integers are in the range -128 to 127. Both can hold 256 different values, but the signed ones are centered around zero.

[@OhsoLosoo]

@@APaleDot clearest explanation I've seen on this in a while.

[@MURDR63]

Couldn't you just use x^1

[@Alexander.Kravchenko]

No, it wil invert nth bit instead of clearing

[@mohanaddarwish2307]

Did anyone got interested about the pile of books behind her and if they are well known or just decoration?

[@aishwaryaramesh4877]

They are indeed well known books! She has authored two of them, the other books are like Bible: CLRS, Algo design from Springer , etc.

[@mohanaddarwish2307]

Aishwarya Ramesh if you are free can you share their names? If you u know them or can read it from the video :)

[@aishwaryaramesh4877]

@Mohanad Darwish , sure! Cracking the Coding Interview, Cracking the tech career , CLRS , Algorithm Design by Springer, some Russian book and I'm not sure about the others... Hope that helps

[@imsubratabiswas]

what she told from 3:38 ? little confusing. can anybody explain ? why to add 1 to both of those ?

[@icono__7136]

You introduced the concept of "inverse" without even defining it. What does inverse mean??

[@Sxhd]

for the twos component you said we just add the inverse so why did you put a 1 if there's already a 1?

[@aysehasan8402]

what about when you shift a negative number to the left ?

[@bradfin12]

Depends on if it is logical or arithmetic. Logical will push the last bit before the sign bit. Arithmetic will multiply the number by two and keep the sign bit.

[@crackcodinginterview4995]

Very clear! Would have been great if it is shown with coding examples.

[@ryzen980]

Atleast the voice in this video is clear

[@SahilRallySuperStar]

Which software are you using to explain things ?

[@johnaweiss]

7:23 I don't understand this notation. (x&(1

[@johnaweiss]

ok, i think `x` is the byte we're evaluating, and `

[@_ashout]

In the last example @8:20, could you not shift a 0 instead of inverting a 1? So (0

[@PepeTostado]

Are there coding challenges (excercises) that deal with bit manipulation? To nail the foundations.

[@jackwindensky5606]

Go to leetcode and filter the problems by bit manipulation, I recommend then sorting by most frequent so you are more likely to find an accompanying video tutorial on youtube if you get stuck.

[@sachiniyer7148]

Thanks.. Yours lectures have been very helpful!!

[@IChowdhury01]

7:20 is the !=0 necessary? Wouldn't just returning the result also work? If it's a 1, 1 != 0 is 1. If it's a 0, 0 != is 0.

[@barisbasar3909]

This lady is lowkey one of the biggest g's in coding lol

[@asawanted]

Regarding clearing the ith bit, can't we do x xor (1

[@Saykey20111]

where did you get that 637. You to explain for those who are not gifted too.

[@ryangriffin5036]

It's a random number that she chose because it nicely demonstrates "carrying the one"

[@namankeshari7332]

The legend herself!!

[@quirkyquester]

how do you flip the bit in python? "~" in python does not flip the bits

[@magnuschase2694]

Hello so when i want to use the arithmeric shift to the right for example on 1011 then i get 1101 and by 0101 i get 0010?

[@sunnyhours84]

Whatta hell?? How ignorant I have been! I've never had to deal with bit operations until now, but I have always presumed that you just use the exact same bits (except the first sign bit) to differentiate between positive or negative numbers. Like say the number 18 in base 10 is 00010010, and -18 is 10010010. Why is this a bad idea, someone?

[@sunnyhours84]

Thanks a bunch for the answer! I've been reading up on different signed representations after my first post, and turns out that some architectures actually do use the "magnitude representation", but they are outdated, so my hunch was not totally wrong. Looks like all modern x86/x84 machines uses "two's complement".

[@jvsnyc]

Not just those, but 19 of 20 others are thereabouts as well.

[@homotopf9461]

@@sunnyhours84 yeah it is not a stupid idea, in your own project you can totally do somethings like that!

[@geraldinejns]

Never mind, got it. It's just an example of how to add up 2 digits. Can't believe i had to watch the video twice to get it.

[@azatakhunov6061]

good video) after watching this i took a certificate at HackerRank

[@bnetjail]

What is "Bitpul Ian Man Tio" ?

[@ankitatrivedi6692]

Bit manipulation :-)

[@antoniotorga5371]

great video LOVING this channel

[@jvsnyc]

This is solid, but I've been somewhat impressed how tricky some of the more advanced bit manipulation problems on HackerRank can get even for someone who's known the basics for years. Sure, you know XOR, but do you REALLY know XOR....

[@trainofthought5733]

how is arithmetic shift of -23 12 i does not work bit wise right??

[@ultimatesin3544]

When is this practical? You know I saw this used in some Google code for a particular object in Android SDK a while back, and it looked like the developer was trying to be clever in squeezing every last bit of efficiency possible out of his/her algorithm.. however if you extended their object in a certain way (which was for a popular use-case at the time) the algorithm then broke for certain situations causing a very elusive bug.. I wish I remember the specifics, think it may have been one of the support libraries.. I do remember it was a pain to track down and I had to override alot of the base methods because of it.. it scared me off of ever using this technique in commercial code..

[@CerealLover56]

Have you never looked at the source code for the java.util packages? The collections defined in that package are going to be used in pretty much every Java application on the planet. They're full of bit operations. They're a fast and efficient way of doing a lot of basic mathematics that would otherwise cause unnecessary slowdown.

[@KingOfAceZ1]

embedded systems? you know, electronics that you can't just download a magical sdk for? you're standing on the shoulders of giants, lol.

[@OhsoLosoo]

@@KingOfAceZ1 alot of CompSci students today don't understand that the stuff they use is built on these very operations that they think no longer serve any practical purpose.

[@ipodtouch470]

@@KingOfAceZ1 I programmed a digital lock in assembly using an 8bit pic microcontroller and you are correct we are indeed standing on the shoulders of giants. The people that are the most guilty are probably web devs, they seem like they have a package for everything...

[@MuhZubairAli]

In binary there is no 2, as being explained 1:38 instead it's 1+1=10 (from this result 0 is placed while 1 is carried to next calculation...

[@jackmocherman1296]

Right, which is why she carries it. Same as in decimal where there is not single symbol for 10. We still describe the condition for carrying a number as two digits in the same place adding up to 10 or more even though there is no "10" symbol, or what is represented as "A" in hexadecimal.

[@FernandoBasso]

The Hackerrank link doesn't work any longer.

[@phil_ka]

that -18 from 18 was damn cool

[@mehdibouchene3208]

how the hell is 8 bit used does the computer have the right to only use 8 at one time or how?

[@geraldinejns]

greate video, but where did you get the numbers 637 and 011.

[@anonymousExposure]

It's a random number that she uses for the purpose of just showing how to add them. So she could have used any other numbers, I got confused by it at first as well.

hope some one help me with my question minute 2:45, i completely understand how we inverse the number, but why arbitrary 10 00 00 00(128)? from e where does it come that?

[@serkandemirel5075]

For example after [phone]) we get -(0) which is 1000000. Although that zero is basically 128 as unsigned 8-bit number, it becomes the first zero before we get to the first negative number, -127 (10000001).

[@proggenius2024]

Great content with good explanation. Thank you👍

[@crackcodinginterview4995]

Good way to explain bits!

[@vamseekotha]

i like the way she says hi

[@MadForCs16]

m a bit overwhelmed after seeing this video

[@muhammadroohallah6091]

Than. You random lady U just explained in a few seconds what my teacher couldn’t over the course of a few lectures (maybe that’s cause i wasn’t listening 🌚)

[@Saykey20111]

1 minute into the video do you know where she got 637? Thanks a lot if you can explain.

[@moj8301]

@@Saykey20111 She was just showing you how to add up numbers

[@jibreelkeddo7030]

Good video but wish it spent more time explaining why we use two's compliment vs one's compliment and the difference. I got stuck on that for a little bit.

[@skilz8098]

The simplest reason that 2's complement is preferred over 1's complement is that with 1's complement you end up with two zeros, a + and a - zero which can lead to ambiguity within the hardware-software implementation of the adders. Also, 2's complement has the effect that when you add any two signed numbers you will get the correct result for both positives and negatives whereas with 1's complement in some situations the result will be off by 1 due to the two zeros that are produced by 1's complement.

[@vikashh384]

Nice Short and Clear.

[@mohamedfouad2304]

What are the books on the table???

[@onionsandwich]

Masterclass !

[@jarrodburns6339]

Excellent video, thank you.

[@Tea-Spin]

How do you tell the difference between -123 and 133?

[@APaleDot]

This is where the distinction between singed and unsigned integers comes in. If the data-type we're talking about is an unsigned 8-bit integer, then 10000101 is 133. If it's a signed 8-bit integer, then 10000101 is -123. Unsigned 8-bit integers can be in the range 0 to 255. Signed 8-bit integers are in the range -128 to 127. Both can hold 256 different values, but the signed ones are centered around zero.

[@sebastiangudino9377]

I'm late, but i want to add that the point of representing signed numbers this way, is that we can use the same adition operation on both signed and unsigned types and still get the correct result. So we can use the same adition circuit to add signed and unsigned integers, we can also implement subtraction by adding A + (-B) and that also works for signed and unsigned numbers, it's a really convenient and powerfull practice

[@johnjonjhonjonathanjohnson3559]

what if a 1 sexually identifies as a 0

[@RickeyBowers]

life gets a bit complicated

[@morguhnburke685]

cant even miss gender numbers anymore can we

[@ShubhamSingh-wm1gh]

one of the best video I've ever seen .. really appreciate your work big thumbs up 👍

[@rajasanchit]

i am forgetting how 11110100 = -12, Can someone explain this?

[@cookergronkberg]

since the left-most bit is 1, you know it is a negative number. To find the value, first take the complement (invert all bits): 00001011, then add 1 Result: 00001100 (12 in binary)

[@ohdude6643]

That "manipulation" square wave got me distracted.

[@emilyhuang2759]

How is logical 0111 0100 = 116 and arithmetic 1111 0100 = -12 ? @6:32

[@protojem]

because, when a number has a 1 in the first bit it indicates a negative number, and a negative number representation is the opposite of the original number, as we saw at 2:15

[@skilz8098]

@@protojem Only when the register type is a signed type! If it's unsigned then there is no negative bit!

[@halo3gamer15]

wait, what did i miss ? Binär: 101 isn´t 583 ?!

[@iloveanime9226]

101 is 1x2^2+0x2^1+1x2^0 = 5 It just like how 105 = 1×10^2+0×10^1+5×10^0 But in binary you have 2 possible number (0 & 1) while in decimal you have 10 possible numbers (0 to 9) and the power is just the index of the number starting at 0 from the right side incrementing as you move to the left so in binary 0 = 0x2^0 = 0 in decimal 1 = 1×2^0 = 1 in d 10 = 1×2^1+0×2^0 = 2 in d 11 = 1×2^1+1×2^1=3 in d 100 = 1×2^2+0×2^1+0×2^0 = 4 in decimal Anything multiplied by 0 is 0 so we can skip that So 100 = 2^2 = 4 583 = 2^9+2^6+2^2+2^1+2^0 So these will be ones and everything in between will be zeros so 583 = 1001000111

[@I_GearV9]

at 0:57 where is Gayle getting 637 from?

[@serkandemirel5075]

It's just a random number that she used to demonstrate how we do addition in base 10.

[@AmirHamedZakeri]

thanks very good explanation

[@techfornoobs4241]

do i really need to know this for tech internships?

[@ITech2005]

Daniel Ro Yes

[@phamanghung2318]

I will fail the interview if asked bitwise operations now lol

[@pwn0x80]

Me to

[@depshallburn]

They don't even teach this shit in my college.

[@huey1153]

Dx D You don’t take discrete math or low level systems?

[@ba1anse]

good video but it is so densely packed, it is not good for beginner to learn at such rapid pace.

[@chefdeprestigellc8855]

Its quick, reduce payback speed

[@plunk6774]

Огромное спасибо за столь хорошее объяснение материала!

[@bhagyasrividyathota6710]

Ma'am please explain how to use constraints in bit manipulation and also with time complexity and space complexity please

[@chiting765]

A little confused at the masks

[@juanhernandez-up4pg]

A bit confused

[@weho_brian]

Gail is a fantastic teacher!

[@sackratte24]

great quality

[@sagunpandit6314]

this is really good.

[@crapadopalese]

Sounds like Rebecca Black is narrating this.

[@wkal1443]

Now my left ear knows about bit manipulation

[@FilipMakaroni_xD]

Oh lord, I've been doing this to "clear a bit" in one of my programs x &= (0

[@sauceslayer5820]

it's a bit confusing

[@AhmadAsmndr]

Thank you very helpful..🌹

[@ian_senior]

uh-ryth-muh-tick > air-ith-matic

[@jasonlee3247]

Poor explanation of 2’s complement

[@sharkwaterproductions77]

I agree, her explanation left me confused.

[@Uniqtech]

Pretty amazing