So coming back here to the anaconda navigator, we can also launch here , jupyter lab. So this now looks a bit more complicated, but actually it isn’t because what you can see here is also a jupyter notebook. So you can always download such a jupyter notebook and import it into normal jupyter notebooks. Or you can also import it into for example, Watson studio or any other cloud hosted jupyter environment. So, they are all compatible and this is also pretty cool and what you can see here is that there are some more windows. So, you see here a little section so this is a file explorer, and then the other thing you see here, you have a number of so call kernel sessions. So, a kernel session is actually more or less attached to a jupyter notebook. So, this means a jupyter notebook is running in the kernel session and the kernel session is executing your actual code. So, you see here this jupyter notebook is of course rendered in my web browser and that’s communicating with that remote kernel. And the kernel is encapsulating the runtime environment, in this case, ordinary python interpreter. You can also shut down those and restart those, here you have some preferences and a list of the open tabs. So the things you can now for example, create a new terminal and you see here there is a second tab opening so you can mix and match jupyter notebooks and your terminals. And you can also drag and drop those around. So you can split and lets add another terminal for example, you can also put it here or here. So it’s a nice way to get and overview of your work. So, create a new python notebook here, and then lets put this here. Ok so you have now, two notebooks and two terminals open at the same time and yeah that’s, pretty cool and that’s basically it. So, now lets close all those. I’ve opened here a notebook from my applied AI course on coursera and you can see here we can execute shell command. So, whenever you see and exclamation mark you can directly execute shell commands, so lets do it here. So this is called a ‘pip install upgrade’ and some packages here so some of those are already installed, but some are being collected. And that’s some sort of a best practice. I’m always doing at the beginning of each notebook I will have a pip install, which make sure that I’m always on the correct versions of the dependent software packages. So you see here the star, star means that the current cell is just actually executed and once the execution finished, you will see a number. And a number is a task ID and that’s a incrementing number. Based on the number you know in which order you have executed the cells. You also see here, that the jupyter notebook is actually running. Now its done, so we see here one and what I always do as well is I’m checking on the correct versions installed. So you see here, I’m checking for tensorflow one fourteen and keras 225. And if that’s not the case I’m raising an exception, so don’t worry about the warnings we are officially on tensorflow 2.x. So I need to upgrade this notebook soon anyway. In case you have a long running task and its just not stopping, you can always restart the kernel. So lets do this here, and I’m saying often restart the kernel and clear all output because, once the output is gone I’m really seeing that the kernel has been restarted. So you see here, all your output is gone and that means you have actually shut down the python interpreter and created a new python interpreter in the background here. Another thing which maybe is interesting, lets actually, lets create a new notebook. So you have here, your output one. You can always say, ‘ collapse all code’ or ‘collapse all output’. And reopen a different cells, that way you can improve visibility of the relevant content. That’s basically it. So there are more things to explore, but I think that knowledge, you can survive. And I think I’ve covered around 80% of what’s important. Maybe one final thing, so you export this notebook into various formats and lets try the ‘reveal.js’ slides. Open it in Firefox immediately, and you see here that’s now a reveal.js presentation.
