NumPy vs Pandas
HTML-код
- Опубликовано: 11 апр 2023
- Data analysis using Python; ibm.biz/Using_Python
Beginner's guide to python; ibm.biz/Python_for_beginners
If you've heard of Pandas and NumPy, you may think one is simply a superset of the other. If so, you're not wrong, but there's more to these two Python-based data analytics packages than that. In this video, Martin Keen explains their relative strengths and ends by offering recommendations on the best approach for adopting one framework versus another.
Get started for free on IBM Cloud → ibm.biz/sign-up-now
Subscribe to see more videos like this in the future → ibm.biz/subscribe-now
#AI #Software #Dev #lightboard #IBM #MartinKeen #Numpy #Pandas #python
Came for the python chat, stayed for the Pink Floyd references
This is undoubtedly the most entertaining and educational video I've encountered on such a complex subject. It cleverly incorporates dad jokes and Pink Floyd references within the initial 60 seconds. I'm absolutely enamoured by it!
I don't think a comparison video is a complex subject. LOL
Thanks for giving us this kind of content, I really love to watch these videos from IBM
amazing explanation! Pls continue to bring such more videos
Thank You And We Need More Like This!
I like this teacher as he explain things very clearly
What's far more amazing than this guys insight into coding languages and concepts, is his seamless ability to write things down on a "window screen" backwards so proficiently. Imagine the pure calligraphy of his "forwards" handwriting
See ibm.biz/write-backwards
This particular IBM presenter is the best!!!
Sir, could you please make a video on understanding YOLO architecture and its various versions or start a playlist on deep learning
I wouldn't mind watching a similar video on seaborn vs matplotlib vs plotly. I find it difficult to choose one of these libraries as my primary goto library. Using all of them is complicated because the syntax and use varies.
Excellent delivery
Very good explanation 😊 thank you!
Well done! Thanks a lot.
Didnt expect a homebrew youtuber to pop up in an IBM vid in my feed 😂
Not very much python libraries related, but I was already thinking Jeff’s videos had R. References, especially Dogs of war (even though that’s from Gilmore I believe…) but now I’m sure there’s Pink Floyd references everywhere :)
Loving the Pink Floyd references, Comfortably NumPy, and Any Colour You Like😁
RUclips pretty much has me pegged.
The homebrew challenge guy chatting about data science libraries.😉
I was waiting for a home brew recipe 😂
Thank you
7,8,9 have special relationship. In positive integers they are the only have the pattern of prime,cubed,squared.
P,P+1=N³,P+2=M²
Why is this pillar of Homebrew RUclips being served to me in a video outlining data analysis via downstream Python programs?
The background of Pandas just change, now is Apache Arrow, but good content by the way.
Sir, how did you write on the screen? How have you positioned yourself while writing on the screen? Seems like you are a leftie but writing direction appears to be right to left ! Great ! Learned a lot. Asked out a curiosity. 😊😊
My guess is, he is probably writing on a huge glass, and the camera is on the other side of the glass. And before uploading, the video is "mirrored" in an editor.
@@LoveYourFamily2 that is correct. It was explained by this channel sometime ago
See ibm.biz/write-backwards
Good explanation
But I want to ask from those who are already related to tech specially ai that which languages are better for an ai developer/researcher instead of python which languages we ought to learn
Java
C++
R
I know python is fundamental and essential
Thanks in advance
Tensorflow (A famous AI (machine learning) framework) uses python for the users to input their data and parameters, but the backend or where the real work is done is written in C++ as it is a faster, more optimized language. So if you were to build your own framework, which I wouldn't recommend, I would use C or C++. However if you are just going to work with the amazing and already available tools then python is your way to go.
Actually it's R , though Java and C++ is best too but it's much harder to use them.
The programing language Julia is quite interesting. It's faster than python, but you can also call python scripts/functions with a package called PyCall. The question about programing languages is always what you wanna do and who is working with you. Most people know Python, so it's the easiest consensus. Are you working alone, but you want to optimize speed on a very complex level C++ is your way to go. Most of the times Python is just enough so I would focus on improving your skills there and if you need more optimization go for something like C++/C# and for analysists R & Julia could be the best ones but Julia did not have it's break through yet.
thanks
please do a video on sklearn, tensorflow, pytorch please!!!!!!!
Amazing explanation. I’ve been crafting my own peogramming « tools », ignoring the libraries available, but that’s definitely the wrong approach as it makes developing a lot more time consuming :/ I’ll learn numpy and then move to pandas if needed :)
Did you do this in production? Or just learning
Wow cool
I am coding a trading bot using numpy custom built function for moving average exponential moving average and min max sliding for float arrays. 😊 numpy is faster
I love Numpy and Pandas. And WOW, for some time we have CUPY (it´s Numpy for GPU, much faster) and POLARS, which is also much faster than Pandas. Try these out!!!
1:00 PYTNOV
Me, running straight to my Jupyter Notebook to find out why 789: 🏃🏃🏃
I feel making a comparassion between NumPy vs. Pandas is wrong, because Pandas is based on NumPy, and it's more to manage data and different kind of documents, meanwhile NumPy it's more mathematical stuff, in a regular day as a data analyst or data science you can use both easily in the same project.
what is this 789 reference?
Pronounced as PythON or PythUN?
What do you mean vs, panda depends on Numpy
Start with numpy! look for the features you are most likely to need and if that answer leads to you to panda………….
Wish "Pandas 2.0 vs Polars" were here.
There are many but understanding simple is much beter unless u are very egur to
Thank you for the suggestion! We'll consider it.
julia vs jax please!
I abandoned PANDAS because it was slow. I wrote my own task specific functions from scratch. But with cython as you have said here, I might have second look.
I'm a daily user of Pandas, but do use Numpy expressly for subroutines to manipulate data to a format I need, as necessary. I find it rare that I will use Numpy on its own.
I wonder how many pink floyd puns he managed to squeeze in
Better is compare numpy || xarray and pandas || polars || PySpark ...
Data science is great and all but have you ever considered home brewing? ;)
why not numpy and pandas ? There are a lot of synergies if you know both
My guy writes better backwards than I do forwards 😅
See ibm.biz/write-backwards
Do a triangle test to see if users can tell the difference 😂 sorry, overlapping interests I guess 😬
Ty, first comment in a row!
Does he have to write everything backwards?
See ibm.biz/write-backwards
@@IBMTechnology ah so simple...and cool
Comfortably Numpy. Lol
Julia is pretty fast
This is way too in touch to be produced by a company
This guy is surely a Pink Floyd fan
Im too
The reason 789 was because needed to eat a 3 squared meal
First comment 😎
You may not know Polars outperform Pandas, and Peaks prepares to outperform Polars.
Ugh, I don't know what's worse... his 789 and comfortably numbpy jokes, or that I found it to be funny. I hate myself.
Pandas was written by people who are very deeply concerned with job security.
Why do you say so?
This guy speaks English but writes doctor.
Good explanation. Sloppy handwriting, though.
Haskell is better!!!!!!
bist cool mit dem spiegelverkehrt schreiben ! NICHT !
what a useless video ... 100% generality, showing absolute NOTHING in detail, no examples of any kind given ...
Comfortably NumPy with the Pink Floyd references