Matplotlib is the only option when there's a lot of data to show. Plotly is often generating SVG charts, and it's very hard to render a very detailed chart composed out of SVG's only. Matplotlib creates raster images and unless you want it to be interactive, it's fine.
plotly is great, but for interactive media. this isn't as good when it comes to publishing static figures in journal articles/books, where seaborn just takes it to another level in terms of formatting and customizability.
@@jannonflores1113 basically power bi and tableau had made this process so simple, that even if this "tool" looks great, it will be easier and faster on the BI tools
I do not recommend using anaconda. Why?: It does not provide a stable programming environment, differ from system to system. What I instead recommend is using the built in venv from python, which gives you more control over your programming environment. It requieres a bit more understanding of the system you are working in and reading more the documents of python, than the plug and play solution from anaconda, but you will earn a proper working transportable environment and a bit more background knowledge.
Wow plotty is so great! I like the way how we can hover on the chart and see the data... I wish I had known this before I completed my graph plotter project which I just finished few days back (referring your matplotlib videos) Anyways Thanks the video was informative!
It would be useful to have a guide on setting up Jupyter notebooks with different kernels. Python is easy to set up - other kernels have never worked for me.
I'm using anaconda and often matplotlib, I know something about plotly and seaborn and I would, I would like to use them more. This video encourages me to do it.
Dumb question but why use or learn any code based platform for serious data visualisation when you can use industry standard dashboards like Tableau, honest question.
at 6:51, how did he copy that line so fast? I didn't even see it all highlight so did he use shift + home or did he use something else to copy that line?
Also discussions about compatibility issues are totally out of date. Conda and pip work flawlessly together, and have done for the last two major releases of conda iirc
Hello can you please do a video on Model training with ADNI 3D MRI dataset. i cant find anywhere. I am trying to train the .nii file but ADNI doesn't provide test dataset. Instead they provide train dataset and test csv. Please make a video on it. Please
Good comparison but what do you mean by saying matplotlib is more professional than plotly? I don't agree on that one and I'm a little confused what you mean by that. Plotly is very professional and has great support and community.
Whatever that takes longer than 15 minutes to put it to work, ain't a "top visualization tool". Tried for first time yesterday, failed miserably to even try the demos... Will return someday, for the time being is in the rubi 🗑️... I mean bucket list 🙂👍
Matplotlib is the only option when there's a lot of data to show. Plotly is often generating SVG charts, and it's very hard to render a very detailed chart composed out of SVG's only. Matplotlib creates raster images and unless you want it to be interactive, it's fine.
Agreed, Plotly is useful for cool visualizations and interaction, but it absolutely dies with large data.
Thanks, that's good to know when you have to choose between the two.
a time saver right here
Wow, I’ve been using matplotlib for so long and never knew how easy plottly is. Guess I’ll try it out!
plotly is great, but for interactive media. this isn't as good when it comes to publishing static figures in journal articles/books, where seaborn just takes it to another level in terms of formatting and customizability.
Yup yup, most of us would just use Power BI or Tableau for this.
@@jannonflores1113 basically power bi and tableau had made this process so simple, that even if this "tool" looks great, it will be easier and faster on the BI tools
I do not recommend using anaconda. Why?: It does not provide a stable programming environment, differ from system to system. What I instead recommend is using the built in venv from python, which gives you more control over your programming environment. It requieres a bit more understanding of the system you are working in and reading more the documents of python, than the plug and play solution from anaconda, but you will earn a proper working transportable environment and a bit more background knowledge.
Jupyter notebook inside VS code is a best environment I think.
Hands on video, just what I need. Thanks man ;)
Really great demonstrations here and well chosen examples.
Wow plotty is so great! I like the way how we can hover on the chart and see the data... I wish I had known this before I completed my graph plotter project which I just finished few days back (referring your matplotlib videos)
Anyways Thanks the video was informative!
Well done! You deserved my subscription!
13:46 wow Turkish life expectancy has been in some weird places.
Florian, thanks for the contents! Every video is really interesting...
It would be useful to have a guide on setting up Jupyter notebooks with different kernels. Python is easy to set up - other kernels have never worked for me.
even on windows, i have used julia and octave calysto kernel just fine with jupyter
That's a really cool lib. Their makeup (blue, white lines) totally reminds me of seaborn
Thanks for the video, very informative!
hätte nie gedacht das du aus österreich kommst :) machst gute videos 👍
I'm using anaconda and often matplotlib, I know something about plotly and seaborn and I would, I would like to use them more. This video encourages me to do it.
Dumb question but why use or learn any code based platform for serious data visualisation when you can use industry standard dashboards like Tableau, honest question.
Costs money
Cool but why are the first 4 minutes a jupyter notebook and conda tutorial?
aconda Tutorial would be a great idea
No
love the new clickbait style thumbnail... worked on me lol!
One word: INTERACTIVE.
at 6:51, how did he copy that line so fast? I didn't even see it all highlight so did he use shift + home or did he use something else to copy that line?
Currently starting development on a visualisation program. I think I'm going for plotly for the interactivity.
How’d it go
@@agonizin went with matplotlib. Couldn't figure out how to integrate plotly with my gui.
@@masherozto study these what ever there in TH-cam is it inhalf to gain knowledge about this or we have to learn more
Yes please do a video on the anaconda environment and why would use conda instead of pip.
Also you are an excellent teacher thank you.
Also discussions about compatibility issues are totally out of date. Conda and pip work flawlessly together, and have done for the last two major releases of conda iirc
I already saw that very jupyter notebook.
Why don't you add a link on it?
Hello can you please do a video on Model training with ADNI 3D MRI dataset. i cant find anywhere. I am trying to train the .nii file but ADNI doesn't provide test dataset. Instead they provide train dataset and test csv. Please make a video on it. Please
Altair is worth looking at... super easy and powerful
I see you doing ci' inside notebooks and can relate so much.
Good comparison but what do you mean by saying matplotlib is more professional than plotly? I don't agree on that one and I'm a little confused what you mean by that. Plotly is very professional and has great support and community.
So the conclusion is that big data uses M, otherwise it uses p, right?
Wie bist du auf die Idee gekommen?
Er hat bestimmt das Video/die Videos von Morpheus gesehen 😂. Es ist schon ziemlich verdächtig, dass beide kurz hintereinander Videos darüber machen.
Great video! Thanks!
I'm trying use matplotlib in an ubuntu server machine.. is there any way that I can view the graphs in server machines?
Jupyter Notebook and a ssh forwarding maybe
Who is this gapminder guy and why does he know so much about Europe?
Wonder if there is any package that manages the Date format of the X-Axis professionally like the way typical stock chart does.
This looks almost the same as seaborn
Jupyter lab is much better than Jupiter notebook.
Like ur thumbnail
Whatever that takes longer than 15 minutes to put it to work, ain't a "top visualization tool".
Tried for first time yesterday, failed miserably to even try the demos...
Will return someday, for the time being is in the rubi 🗑️... I mean bucket list 🙂👍
Are you talking about plotly? Can't imagine how setting it up would take you anywhere close to 15mins.
awesome
I don't like Seaborn, can't handle too much data on the pairplot
Yes, it is.
Gosh, you spent so long waffling before saying anything about plotly.
For sure far better
👏👏👏👏👏👏
First comment! I really like your videos
yes and no at the same time
Wow, so much japanese here, it's simply mindboggling.
👌👌👌
Is plotly the better matplotlib? Hell no
Matplotlib is still the best for me
HELP
funny face 😂😂