Set Up a TensorFlow Development Environment

0:00Hello and welcome. My name is Jonathan Barrios and in this skill we're going to

0:04create a TensorFlow development environment. Let's check out everything

0:09that we're going to learn in this skill. The first thing we're going to jump

0:13into

0:14is Google's Colab and Google's Colab 1L is a or an online version of Jupiter

0:27notebook. So what is Google's Colab and what is Jupiter notebook? Well these

0:34are

0:34iterative development environments where you write code in a cell and then

0:40when you execute that cell you get the output right away and you can have as

0:44many cells as you like and then you write some more code and you get output and

0:48you can also get some visualizations of graphs and things like that. When you

0:52're

0:52working with an IDE it's going to be a little different. You're going to have

0:57all your code in one file and then here you would have it like a terminal and

1:03then you can execute that code and then you would get the readings of that code

1:07out here. So it's just a different workflow. One of the cool things about

1:12Google's Colab is that it has everything that you need for data science and

1:17machine learning. For example the packages and libraries already installed

1:21for you whereas Jupiter notebook is a local version so I'm going to put local

1:25and so you'd have to install those yourselves. This is on the cloud made

1:30available by Google. So number two the second video is we're going to check out

1:35something called anaconda that comes with something called conduct. So these

1:41aren't snakes they're snake names but this is really a Python distribution

1:48which

1:48includes all of the ML and data science packages that you could need for pretty

1:54much anything you can think of. And if you need to add something custom you can

1:59just simply add it using conduct. And then number three we're actually going to

2:05check out Jupiter notebook. And we're going to try to recreate what Colab does.

2:11So we're going to take two and three and try to equal the awesome star power of

2:17Google's Colab meaning that we can use a local development environment instead

2:23of

2:23a cloud environment but get almost the same resources. The only thing that to

2:27keep in mind is when you're using Google's Colab they offer GPUs for free. So

2:34you're

2:35not going to get a free GPU at home in a local environment unless you have some

2:39sort of a hookup. So just keep that in mind that's a pretty big distinction.

2:45And

2:46then for number four we're going to jump into IDEs and we're going to check out

2:50PyCharm which is an IDE by a company called JetBrains. And then number five

2:59we're going to check out VS code which is by Microsoft. And funny enough this

3:08is

3:08open source and free to use. And it has a lot of extensions so you can extend

3:16it.

3:16And this is probably what we're going to end up using most of the time but we

3:22will be using almost all of these tools at different times. And then number six

3:28we'll have a challenge where you get to actually apply everything that you

3:32learned in this skill to some sort of final output. In this case it's going to

3:38be some kind of TensorFlow development environment. So what is the context here

3:44?

3:44Well so far in the introduction for machine learning and introduction for

3:49deep learning courses we've used Google's CoLab. It's a great way to get

3:54started because you don't really have to worry about a development environment

3:57and you can just start working. There are some headaches because it is a

4:03computer somewhere in the data center somewhere. So when you want to transfer

4:08files forward and backwards it's not as easy as a local environment. So it's

4:13because it's a computer somewhere else. So there's some hurdles to jump through

4:16.

4:16That's just something to keep in mind. The main power is the ability to use the

4:21GPU in Google CoLab. So when you're using a local development environment it's

4:26good to remember that you're going to be responsible for the GPU and the

4:30compute and all of those things and storage. So that's a huge trade-off.

4:34We'll be discussing this as we go through this course. And finally at the very

4:38end when we get to the challenge you're going to build your own development

4:41environment of choice and well actually not of choice. It's going to be visual

4:46studio code but you're going to do something really specific to set up this

4:50development environment. Then with all the skills that you have you can set up

4:55a development environment of your choice. I guess that's kind of what I was

4:59getting to. And the challenge is VS code but after that you can do pretty much

5:03any of these combinations that you find advantages for your project. All right

5:09so I will see you in the next video which is going to be that video on Google

5:12CoLab. See you there.

0:00Welcome back. In this video we're going to check out Google Colab but first I

0:04wanted to share this repo with you. This CBT nuggets github and this repo that

0:12I

0:12wanted to show you is here Introduction to Deep Learning. We're going to have

0:15our

0:15own repo soon but first I wanted to show you this sunspots CSV right here and I

0:22'm

0:22going to click on RAW so that I can show you how Google Colab works and I'm

0:27just

0:27going to copy this RAW CSV. Now I'm going to go over here to Colab and I search

0:32for

0:32Colab in Google 1L1 word and I'm just going to click on this first one which is

0:38colab.research.google.com and once you're in there and if it's your first time

0:43just

0:43go ahead and click on new notebook and this is going to open up Google Colab.

0:48You definitely have to be signed in to a Google account and that's all you

0:52really

0:52need. Generate with AI here that's something new and I would suggest not

1:01using that and just keep in mind that if you do use it it's going to be a trial

1:05basis and you're going to have to pay for Colab Pro basically but it's a good

1:09idea

1:09when you're first starting out and you're really trying to learn TensorFlow not

1:13to

1:13use AI to write everything yourself. If you're going to use AI ask it questions

1:19to help you learn. So you use it as a tutor but never just copy and paste

1:24anything. Even don't copy and paste the code that I give you. Just code along

1:29with me because that helps you get into long-term memory so that you can write

1:32the code yourself. I just wanted to just give you a heads up about that AI and

1:37now let's do some stuff. Okay so what can we do? The first thing I want to do

1:42is

1:42show you that you can talk to the terminal, the command line, by using the

1:46exclamation mark. I'm going to say Python - version and let this run and to run

1:52this I'm going to shift enter and while this is running this is the very first

1:57time you see that it's connected to a computer somewhere in the cloud. Now it's

2:01connected and we see that we're running Python 3.10. Alright so now I'm going

2:06to

2:06show you how you can import things such as pandas and you don't need basically

2:13any pip install. So let me explain that. Normally when you are working with

2:18packages locally you have to install them. Right so you like let's say you're

2:22using like a Python package manager like pip you'd have to say pip install pand

2:29as.

2:29If you're using Conda environment manager then with Anaconda you would use

2:35Conda install pandas so on. So here this is done for you and it's really nice.

2:40When

2:40we get to Jupyter Notebook which is a local environment we're going to have

2:44to install these using Conda. You could also install it with pip so you have a

2:47choice. So I just wanted to give you a heads up that the nice thing about Colab

2:51is that has all of this included. So if I run this we're not going to get any

2:57errors and we can actually define a data frame and I'm going to say df equals

3:01pd.read_csv and I'm going to pass in that raw file and then I'm going to run

3:08that and then I'm going to call the first three rows of that data frame so we

3:15can

3:15just take a look at it. And now you can see that we're just actually getting

3:19ready to start some data analysis or data science or machine learning right out

3:24of the box. But I think the most important distinction is if you go up

3:29here to runtime and then go to change runtime type you can go from the CPU

3:35to a GPU and I'm going to say okay let's save and now it's connecting to a GPU

3:42and we're there. So look at that we have a Tesla GPU

3:46and it's free. This is pretty amazing because

3:50GPUs are not cheap especially right now in 2024 they're getting pretty

3:54expensive.

3:54Every year is going to get more and more expensive or it's going to become very

3:59scarce because more and more companies and individuals are going to start to

4:03use these resources. So when you're working locally with

4:08Jupyter Notebook you need to have some sort of a GPU. An external GPU card

4:13is very common for PCs but very currently Apple Silicon has made GPUs

4:20in their own chip. So I'm using the M3 chip and that's actually able to run

4:29a lot of models on it locally without having an external GPU. So that's

4:33something new we'll talk more about that but this is something that's really

4:37crucial to keep in mind if you don't have a GPU you might need to use something

4:41like Google Colab. In fact a lot of engineers at

4:44Colab, excuse me at Google, would use Chromebooks and then log into

4:49computers in the back end. So that's how they're able to use Chromebooks

4:52because

4:53Chromebooks don't have CPU or RAM or very much storage at all.

4:58So that's kind of something to keep in mind and it's also a trend.

5:03And we're not doing anything to use a GPU so I'm going to show you the best

5:06practice just to go up here disconnect and delete runtime.

5:11That's how you do that or you can just go back to the CPU by going to

5:14change runtime type. All right so let me close this resources here

5:18gives you a little bit more information about this and again another

5:21opportunity

5:22to upgrade to Colab Pro.

5:26So that's it for this super brief introduction to Google Colab. We'll be

5:30seeing our friend again but for right now we're going to jump into

5:33Anaconda and Conda. See you there.

0:00Welcome back. In this video, we're going to check out Anaconda, not the snake,

0:05but the Python distribution.

0:07Alright, so I just typed in Anaconda and I'm going to click on this first one

0:13right here, "Unleash AI Innovation and Value."

0:16That's very corporate sounding technique. Alright, so basically, let's just

0:21skip all of this stuff and go to the free download.

0:24We're going to skip registration and then here we can see that we have the

0:28different options.

0:30So if you're on Windows is pretty straightforward. You only have one option. It

0:34's a graphical installer.

0:35With Mac, you can use homebrew and you have these other options, but what I

0:38would recommend, I'm on a silicon, so I use this.

0:42But if you are not and you're using Intel or everything else, just use this one

0:45.

0:46And if you're on Linux, you'll know what to use.

0:49Alright, so I've already installed this, so I'm just going to skip this stage

0:54and then once you have installed this, then unpause or just come back and join

0:59me.

1:00Once you're done installing and it's actually, this usually pops up by itself.

1:06So this would be the first thing you see.

1:08If you don't see this and it's not popped up, then just click on Anaconda Nav

1:11igator.

1:12And here you can see that we have Jupyter Notebook. We can just click on launch

1:16and it would work straight away.

1:18And what else do we have? We have Anaconda Notebook, so they're doing their own

1:23cloud version of notebooks.

1:25And here we have PyCharm. This is what we're going to talk about as well.

1:28We're not going to get into Spider, but that's another IDE.

1:31And Visual Studio Code, which we're going to talk about, it's used to be here,

1:35but it's not here any longer.

1:37I could just click on this button here to launch Jupyter Notebooks, but we're

1:42going to do that in a separate video and we're going to do it a different way.

1:45We're not going to just rely on the Navigator to open it up and click this

1:48every time we want to run this.

1:51And instead we're going to go over here and work with this a little bit. Let me

1:55make this a little bit bigger.

1:57Alright, now that we're here, you can see that we don't have, it doesn't say

2:03base here, but if it did,

2:06it should if you've installed Anaconda. So now I'm going to say Conda Activate.

2:12Then you'll see this. This is probably what yours has already, if you're on a

2:19Mac.

2:20Now if you want to find this for Windows, it's called the Anaconda Prompt.

2:25And what you want to do is just go to the Start menu and press on the Windows

2:29key and then type Anaconda Prompt into the search bar.

2:34And when you do that, you're going to probably see the Anaconda Prompt pop up.

2:38If you don't find it there, it's also available in the Start menu by scrolling

2:43through the list of apps under the Start menu.

2:46Alright, so once you have that up and running, I have Conda activated here, but

2:51this is a base environment where we have all of the packages involved like

2:55everything.

2:56So you don't want to use this. And instead what you want to do is create an

2:59environment.

3:00This is how you do that. Conda create dash dash name and then the name of the

3:07environment.

3:08I'm going to call mine TF_PRO. And then you would install whatever you needed.

3:14Let's say pandas, right, and TensorFlow.

3:19And then once you do this, you click on this and then you scroll through the,

3:23basically it'll ask you yes or no if you want to create it and you say yes,

3:27and it'll create this for you and it'll also install pandas and TensorFlow.

3:32And then once you do that, it'll activate that base for you and you'll see that

3:36here.

3:37I've already created this and so I'm not going to create it, but that's how you

3:40would do that if you wanted to create your own new environment.

3:44And you might be asking, well, why are we doing this?

3:48The reason is you might have one project that has web development stuff like

3:53Python and Django.

3:55You might have another project has nothing to do with web development and it's

3:58just machine learning.

4:00So you have TensorFlow and pandas. The thing about this is when you share your

4:05projects, you don't want to share every project or every package ever made with

4:10every project.

4:11So you want to keep it as tight as possible.

4:14So we only install the things that we need for that project.

4:17And that's very efficient. So you can imagine if you're sharing this with

4:20people, you want them to make it as simple as possible.

4:24So that's some of the reasons why.

4:27All right, so let me keep moving on.

4:30And now what I'm going to do is say, oh, I forgot the name of my environment.

4:35So I'm going to say, Conda and list so that it can show me.

4:38And I can see, oh, here it is. Tf Pro. That's the one that I created.

4:42So I'm going to say Conda activate Tf underscore.

4:48Oops.

4:50Tf underscore pro. There we go.

4:54And now you can see that we're activated. And this is going to be the project

4:58that I'm going to use for the professional TensorFlow developer course.

5:04Right. And so this is why I would call it this. And then I would have all the

5:07packages needed for that project.

5:10If I wanted to run Jupyter Notebook, I would have to install it. So I would say

5:14Conda install.

5:16And then Jupyter, which I've already installed.

5:20But this is how you would do that.

5:22And if you wanted to run it, then all you would have to do is just type in Jupy

5:26ter Notebook.

5:28From wherever your project is. So we're going to do that in the next video.

5:33But I'm just giving you a heads up that that's how it works. And let's say we

5:35're done now.

5:37Okay, I'm done developing. So I'm going to say Conda deactivate.

5:41And when you do that, it goes from TF underscore pro to the base.

5:45But if I did it again, then it just goes away.

5:49All right. So that's about it for this video. And I will see you in the next

5:52video when we jump into Jupyter Notebooks.

0:00Welcome back. In this video we're going to dive right into Jupyter Notebook

0:04from

0:04the terminal right where we left off of. So now here how do we get started?

0:09Well

0:09first of all I'm going to activate Conda. So Conda activate and then what we're

0:16going to do is Conda activate and then the name of the project so it's TF_PRO.

0:22Okay and let's just say that I'm not in that I'm in my users folder but I want

0:31to

0:31be in my project folder. So let me just create one here and I'm just going to

0:35name this github. So what I'm going to do is navigate to das tab and it auto

0:45completes for me and then I'm going to say gi tab and then at auto completes it

0:50kind of knows what these names are in these directories just a little tip there

0:54and then if I hit enter now I'm in that github folder on the desktop. So from

1:00here I'm going to say Jupyter Notebook because I've already installed this

1:06using Conda install Jupyter. So now that I have that installed in this

1:11environment I can just do this and instantly it creates this environment and

1:20I'm going to go over here to new notebook. There's our notebook now I can

1:25double

1:25click on that and now we get to choose a kernel. We're not going to get too

1:31deep

1:31into this right now. If you click on this you can see there's no additional

1:34kernels. There's a way to deal with this so what I'm going to do for right now

1:38is

1:38just select a Python 3iPy kernel. Okay so now that I'm in here I've already

1:44installed Pandas right so if I said let's try the Python - - version just like

1:52we

1:53didn't call that and we're using a newer version of Python 3.12. Okay and what

1:58else? Let's try some Pandas so let's import Pandas. It's PD and I'm going to

2:06define this URL that we copied earlier like this and then gf I'm going to

2:13create

2:13a data frame and use PD.read_csv and then I'm just going to pass in the URL

2:22like that and then I'm going to call the first three the head three first three

2:29rows. Alright let's try that and instantly we get it access that's the

2:36repo grab that CSV read the first three and we're ready to rock so this is

2:40Jupiter notebooks and the only caveat here is that you can't go to kernel and

2:46say give me a GPU because that's your job so we're responsible in a local

2:52environment for providing the GPU in any of the compute requirements something

2:57to

2:57keep in mind and makes these online computers really attractive right so

3:03there's a lot of services that offer this so just keep that in mind but as you

3:09see here it says shut down kernel shut down all kernels what's all this well

3:13this has a little bit to do with how we want to get out of this so there's

3:17different ways of closing this down if you wanted to stop you could interrupt

3:22a kernel like this but what it's pretty common is to go here to file and then

3:28go to shutdown or you can close and shut down notebooks let's try that and

3:35yes I'm okay with that alright there it is and then now we're here if we go to

3:40running nothing is running and then we can go to file and then shut down and

3:47let's shut that down and if we go back over here you can see that our server is

3:52no longer running but we're still here in TF Pro and if you wanted to you could

3:57deactivate conduct deactivate and now we're in base and if you want to do that

4:06again now we're completely out and we can type clear and that's it that's the

4:10end of our session all right so now we're ready for IDEs and I will see you the

4:16next video where we jump into PyCharm and VS code see you there

0:00Welcome back. In this video, we're going to check out IDEs because we've been

0:05looking at Google Colab and Jupyter Notebook and those are interactive

0:10notebooks. Right? They have these code cells, you write something and then you

0:13get

0:14that output, which is really nice. When you're working with an IDE, you're

0:17usually

0:18working with a .py file if you're working with Python and then you're going to

0:21execute that script in a command line somewhere and then it'll give you an

0:26output and it's really hard to see visualizations and things like that. It

0:30can happen, but it's really a clunky way to do it in an IDE. Things have gotten

0:35a

0:35lot better and that's why you basically can work with an IDE now. But for a

0:40minute

0:40there, it was kind of like there was people that were using interactive

0:43notebooks and there's people that were using IDEs. Now the thing happens

0:48where you get this interactive notebook like it could be Colab or it can be

0:55Jupyter Notebooks, that file extension is .ipynb and that stands for IPython

1:02notebook. So that used to be the name of Jupyter Notebook before and then it

1:09changed the names to Jupyter Notebook. So instead of IPython Notebook, we have

1:12Jupyter Notebook but it keeps that file extension anyway. Long story short that

1:16IPynb file extension for production usually needs to get converted to a .py.py

1:22file and so that's why working in an IDE is one of the reasons at least that

1:27makes it very useful. So let's dive into it right now. However, we're not going

1:32to

1:32be using PyCharm. We're just going to kind of talk about it but we're going to

1:36spend most of the time working with Visual Studio Code because first of all

1:40it's free and it has a ton of extensions so we can make it very very flexible.

1:46Alright, let's do this. So here we're going to make your software reality, make

1:51your

1:51software vision a reality. Alright, so developer tools and where is our reality

1:56?

1:56PyCharm. Okay, so here it says the Python IDE for data science and web

2:02development. So that's a lot. So this is like an 18-wheeler compared to driving

2:11your comfortable Honda, right? So that's a famous saying by the person that

2:18created Python and it's true, right? They were talking about different IDEs but

2:24it's a very apt comment because this IDE can do so much stuff and you might not

2:30need to do a lot of that stuff. For example, if you don't do anything else

2:33except for data science, you're not going to need to do web development. So all

2:37of

2:37that, which is a huge amount of tools, can just go away. So it's something to

2:42consider. Alright, so let's take a look at this and this is what it looks like.

2:47Pretty much looks like Visual Studio Code if you're familiar with that

2:51interface

2:52and definitely check out this video if you want to. So here you can see that we

2:57have Python which is great but we also have Django, Flask and Fast API which

3:01are

3:01mostly, I would say these two are for web development. Jupiter, SQL, Anaconda,

3:06and

3:06Pandas, well this is kind of more our sort of flavor.

3:12Alright, so it empowers developers to be more productive and makes difficult

3:15coding tasks easier. And yeah, so again, it's a little bit overkill. There's a

3:22debugger, all this great stuff. Now if you want to dive into the data science

3:29part, we can just click on this to sort of focus in on that and it

3:33can go into all this stuff and this context aware AI, again, that's something

3:37that we saw even with Google Colab and so everybody's going to be doing this

3:42and here you can see that wow, everything that you need, even hugging face,

3:46TensorFlow, Keras, scikit, learn. Anaconda, but it's right and even Docker, we

3:53're

3:53not talking about Docker but this is something that's very, also very, very

3:56relevant. So if you wanted to download this, let's go to the download site and

4:01it'll basically see that I'm on a Mac and then just say download this one.

4:05If you're in Windows, it's also a downloader and we're okay. So basically we

4:10don't

4:11have a lot of options here. Let me see for Intel. Okay, cool. We do have an

4:15Apple Silicon. This is what I would be downloading. So just, you know, if you

4:19want to do this, download it and check it out, you can do that or you can go to

4:23an anaconda navigator and where's PyCharm? PyCharm Professional Install, right?

4:30So this is a full-fledged IDE, meaning that 18 Wheeler, that's why we're

4:34talking

4:34about does a lot of stuff here's talking mostly about web development.

4:39All right, so that's it for this video and I will see you in the next video

4:42when we

4:43jump into Visual Studio Code. See you there.

0:00Welcome back. Alright, let's dive into Visual Studio Code by downloading it.

0:05And here is the download and you have some different options. So for me there's

0:10no silicon so I would just basically use the stable universal installer. Once

0:18you

0:18have that installed we can go over here to our desktop and we can try to create

0:23some projects here. So this is what we used before so I'm just going to exit

0:29and

0:30get a new one and I'm going to make this a little bit bigger and what I'd like

0:35to

0:35do now is create that project. So I'm going to Conda activate. Alright, so we

0:40new Conda and a Conda install and so now we have our base loader here and we

0:46can

0:46create some environments or some yeah some Conda environments. So I'm not sure

0:53if

0:53you installed it before but if you did great if not let's do that now. So we

0:58want to say Conda create and then dash dash name. Then before I finish this I

1:05do

1:05want to share this Conda cheat sheet with you and this will be linked in the

1:09skill and here you can see basically anything that you need to do. So here is

1:14Conda create dash dash name and then this is the name of the environment and

1:18then

1:18here you can install Python whatever version you want or if you leave this

1:23blanket will install the newest one the latest version that it has and then you

1:28can install other things other packages right. So if you wanted to install

1:32pandas you think you can do that here as well. So this is a really great cheat

1:39to have like Conda and list is right here Conda list will list all the packages

1:44and versions installed in that active environment whatever is in the

1:48parentheses.

1:49Okay, so we're going to create name and then we're going to say TF underscore

1:54pro which I already created but I'm just showing you how you would do that and

1:58what we want to do here is add pandas and TensorFlow and we're also going to

2:05add

2:05Jupiter like this and run that and then you'll have you can activate that by

2:11using this name so Conda activate TF underscore pro. Now let's say that you've

2:17already done that but you've and you only installed it pandas and TensorFlow

2:21but

2:21you did not install the others so what you could do there is first Conda

2:26activate

2:26and it's TF underscore pro and now that you're here if you need to install

2:35Jupiter you can say Conda install Jupiter like that and then once you do that

2:43then

2:43you can run Jupiter notebooks but we're more interested in Visual Studio Code

2:50so

2:50what we're going to do here is just make sure that we have this TF underscore

2:54pro created and have all the packages installed that we're going to use we

2:59can add more later but we have this done so now I'm going to say Conda de

3:03activate

3:05I'll do that twice all right and then clear and then boom we have a clear

3:10terminal and now we're ready to start working with Visual Studio Code. All

3:17right

3:17so this is a new window and I've already created a project but we're not going

3:22to look at that just yet let's just think of this as the first time that we've

3:25run this if you use shift command P you can get this sort of area and then you

3:31can add some different things for example you can say execute code shell

3:42command

3:42install code in the command path that's something that you want to install so

3:46you could write it like however you want to write it to get to this point so if

3:51you click on this it's going to install code in the command path and let me

3:54show you what that does once you have that done you can go ahead and close

3:59Visual Studio Code I don't even hit quit here and then what I want to do is I

4:04want

4:04to navigate to cd desktop and then inside of that desktop I want to go to

4:11TF underscore pro let's go in there now we're there and now from here I can say

4:18something like code and then period when I do that it's going to open that oh

4:25it's

4:25not working oh I know why let's open this up and install it together because I

4:31have not installed that I thought I did so it's run code or is it once it

4:41install code from there it is so shell command I'm going to click on it and now

4:49it's going to say okay and then my password all right so shell command

4:57successfully installed in path and so now I'm going to close this and then I'm

5:02going to try that again because I'm still in the terminal and I am in that

5:07correct directory so if I use code period it's going to open that up in that

5:14directory so that's something really handy to have so that's the first thing

5:18all right so before we get into anything else I want to show you what all of

5:22this

5:22is here on the side okay so first this is the file explorer and you can see

5:27here

5:27that we have the different file types and this is where you would create new

5:33folders and such and then here this is searched and you can search all kinds of

5:38things in here for missing files or certain classes and things like that and

5:43source control is really cool and this is where you could connect to you get

5:48hub repository for example and push your changes to that repository which will

5:53we will be doing quite a bit of and then here is a debugger so you can actually

5:57run this to debug this is mostly something that you would find a lot in

6:02web development and those kinds of workflows but it's also you would see this

6:06in data science as well it's very useful and so this is a debugger and then

6:10here

6:10are some extensions so let's go ahead and install some extensions now so I'm

6:16going to say Python ext for this extension so what I want to do here is

6:23install this one and this one and this one so these three that's it so don't

6:31install this one this one or this one that's my recommendation and we don't

6:36need a python environment manager because we're using a conda okay so once

6:42you install these three we can move on to the next one which is Jupiter there

6:50we go click on this and I've already installed this you can see that is

6:55installed so go ahead and install this and then once you do that we'll get into

7:01how you can actually have the best of both worlds meaning you can have

7:07something

7:07that looks just like an interactive notebook right inside of visual studio

7:11code and there's multiple ways that you can actually leverage this because it's

7:16really useful to have the iterative code cells so that you can see things run

7:22as

7:22you're working because it's common to jump back and forth right but if you're

7:26in a dot pi file it's actually not that great because you need to run the whole

7:31script but this plugin will allow us to like highlight a section and then see

7:37it

7:38run in like a virtual notebook so it's like pretty amazing stuff so let's okay

7:44definitely install this let's keep moving there's so much to install that

7:47there's all kinds of stuff so I'm gonna say dark theme okay and this is the one

7:57that I usually use and so I'm going to go ahead and install this one and once

8:04you do that you can go over here to settings and profiles not themes and I'm

8:16gonna say color theme and here you can set them up so I'm gonna say dark plus

8:22okay and then another cool thing that you can do is set up some sort of icon

8:27theme I use this material so I'm gonna install that one okay and then you can

8:35set it here or you can go down here to themes and file icon theme that's the

8:41one that I want and we're gonna use the material icon well we can use either

8:51one of these let's just use this one okay and what does that do well here it

8:56actually changes these little icons to look really cool and this is I find it's

9:00really useful to tell the difference between files and so I really like that

9:04and I like this dark theme too all right so there's really a lot more that we

9:10can install here but I mean just just a lot so take a look but I wouldn't

9:15install too many more because you can actually get some conflicts and that's

9:19actually I'm glad I said that because we go here and we go to this read me I

9:25wanted to bring this to your attention so let me show you how you can get the

9:29terminal click up here and do terminal or you can use this shortcut and here

9:34you can see that there's no base or anything like that so how do you deal

9:41with that so down here well this is a read me file but I wanted to show you

9:46that if you activate it if you said come to activate it would say base and then

9:50you can activate TF_pro or whatever name of your environment the problem here

9:57though is that it would say base and the TF_ so it was showing too so this is I

10:04went to stack overflow here let me show you this link there we go so you can

10:12see

10:12it says base and then whatever the name of that environment is so that was

10:16happening and so that can happen with conflicts so if you go down here the

10:22reason that there was like there we go as described on this page the Python

10:27extension now automatically activates the environment so it seems that the

10:32problem is caused by the automatic activation it's by that plugin so you

10:37could use this right here to copy this and basically go back to the terminal

10:42and

10:42then paste it there run it and then restart this sometimes you have to

10:45restart the computer but then it will fix that problem so that is something

10:50that can happen and this is just read me file .md for markdown where you can

10:55just write text in it all right so let me show you something really cool I'm

10:59gonna get out of here and I'm actually just going to close this just quit code

11:03completely and then I'm gonna open this folder that I have here this is the

11:08folder that we've been opening up in Visual Studio code you could just say

11:12code period because we're in that folder here in the terminal or you can grab

11:17it

11:18and just drag it in which is not the most programmatic way of doing it but you

11:23could do it that way or go back to this folder and you see that I have this

11:31workspace file if I double click on this look what happens this is really key

11:38right here you didn't really notice this but out but here is the Kanda

11:43environment which is this is pretty important this is kind of tricky let me

11:49show you let's open up the terminal okay if I click on this it shows you that

11:54it's

11:54a Kanda environment that we created it's not the base but the pro and this is

11:58exactly what we want to have if you go to terminal and go to new terminal you

12:03can see that we have that selected so this is pretty cool this is how you use

12:07these Kanda environments in the right kernel using VS code this is pretty cool

12:13and don't worry if this is a little bit overwhelming it's a lot of information

12:17we're gonna go over this again and again and again and so every time we start a

12:23skill we're going to set up our development environment you know using

12:27whatever we're using but it's going to be a review of this so it's going to be

12:30something that's not just like all this information all at once all right so

12:34what we're gonna do now in the next video is set up our challenge so I'm

12:38gonna try to get you to to to accept this challenge and make it interesting

12:44for you and well the main goal is to create a main.py file a readme.md and

12:50then once you do that you want to go to file and then save workspace as and

12:59once

13:00you do that you'll have this workspace safe and then I'm going to show you in

13:03that video the solution video how to get the proper environment and if you've

13:10already created that your own con environment is just by clicking this and

13:13selecting it really so if you do that and then you save your workspace every

13:18time you click on that workspace file it'll open this up and you'll be in the

13:22proper con environment you don't have to write anything in the terminal pretty

13:27awesome so I'll see you in that next video where we kick off the challenge

0:00Welcome back and congratulations on making it to the end of the skill and to

0:04the challenge.

0:06This is going to be awesome.

0:07Let's do this.

0:08All right.

0:09So here we have this empty page.

0:12And what we want to do in this is add some code.

0:15But first let's close this and your challenge is to create your own Kanda

0:22environment, a

0:23new one and then a new folder for your project.

0:28To navigate to that folder in the terminal and use code space period to open it

0:35.

0:35So mine is TF_PRO.

0:37Yours can be whatever you name it and then open that project using this method.

0:41That's part of the challenge.

0:42And once you do, let me actually just go ahead and do that.

0:46And once you do, you want to create a readme.

0:48And in here you can say anything like hello or like I did it or whatever you

0:53want.

0:54And then here in the main.py you're going to do what we did before in previous

0:59videos.

1:00And I'm just going to make this pretty straightforward.

1:02So you're going to add a comment here.

1:07So import pandas with alias.

1:13So I'm going to leave that as a trick.

1:15Not a trick but a challenge.

1:18And then here what you want to do is create a data frame.

1:23So this should be pretty easy but I'm going to make it even easier.

1:26So it's PD.read_csv and then in here you're going to pass in the URL.

1:36And you're going to have to define that URL up here and put it in quotes.

1:42And then you can uncomment all of this stuff and then just write the code down

1:46here.

1:46This is just comments to help you out.

1:49And then in the next video we're going to, in the solution video, we're going

1:52to, I'm

1:52going to show you how to do all of this stuff from scratch.

1:56And make sure that you are in the correct conduct environment that you created.

2:01Right?

2:02You don't want to be in base.

2:03And if you want to double check you can go to terminal and say new terminal to

2:06see if

2:06you are really in there.

2:08All right.

2:09So you got this and I will see you in the solution video next.

0:00Congrats on completing the challenge and let's take a look.

0:04So first things that we want to do here is navigate to the right directory and

0:09I can use

0:10pwd print working directory to see that I'm already in the right directory.

0:14So that's cool.

0:15So now I'm going to say code space period to open Visual Studio code in that

0:18directory

0:19and now look where we are.

0:21So let's just say that I created this and the way that I would do that is click

0:24on this

0:24plus sign and just say main.py.

0:28But it already exists.

0:29I'm not going to do that.

0:30I want to create the read me.

0:31I would also do that all caps.

0:33That's the convention read me dot MD for markdown.

0:37And then this is what we're going to do last.

0:40All right.

0:42So here let's do the first thing.

0:43So we need to import pandas with the correct alias.

0:46So import pandas as pd and we're going to define that URL here.

0:57So let me go to this place introduction to deep learning sunspots sunspot CSV

1:07grab that.

1:08Oh, I need raw first because otherwise it will not work.

1:11It needs to look like this.

1:12Then we grab that URL go back put it between the double quotes.

1:19All right.

1:20So now that we have pandas imported, we have the URL.

1:23We can say df equals pd dot read underscore CSV and just pass in that URL.

1:32And then we're going to say D we have to use print and I'll explain why print.

1:38And then here we're going to say data frame dot head.

1:44And let's just look at the first three.

1:46If you want to look at the last three, we can do that using tail.

1:50Let's not trail tail.

1:53All right.

1:54Let's save this and then now what we're going to do is execute this.

1:57I'm going to say main Python main dot hi.

2:06And there's the last three.

2:08All right.

2:09So this is how you complete the challenge except for one last thing.

2:12All right.

2:13So now that we've done this and congrats for being able to do all this is

2:16really awesome.

2:17So now what I'm going to do is I've got this set up and I'm going to make sure

2:21I'm going

2:21to go over here.

2:24I do see that we're using the right TF underscore pro conduct environment.

2:31So that's good.

2:32So now that we have it like this, we don't want to have to set it every time

2:36because

2:36you do have to click on this every single time.

2:39It's kind of a pain.

2:40So the way we work around that is now say save workspace as and then I would

2:45basically

2:45save that in whatever you finally want and then click save.

2:50I've already done that.

2:52Let me show you what the benefit of doing that is the benefit is this whenever

2:55you want

2:55to start working on your project, you just go to your project, you open up that

2:59folder

2:59and you double click on this and you're done.

3:03And you can see that you have the TF pro already selected and you're ready to

3:08go.

3:08If you wanted to run that file again, you could, but we're not going to do that

3:12.

3:13All right.

3:14So that's it for this video.

3:16And wow, we've covered a lot of information and don't worry if it's not all 100

3:21% clear.

3:21We're going to be using all of these tools in one way or another because we

3:26have to use

3:26different tools and that's what we're going to be doing.

3:29So until next time, I hope this has been informative and like, thank you for

3:32viewing.