A few months ago, I tried a free trial of Team Treehouse and then signed up for one of their basic subscriptions. The subscription let you enroll into a learning track of your choice, and I wanted to start with Python. The Beginning Python track is 9 hours long, but it took me a few months to complete. However, today–right now, I have finally passed a major milestone in my Python learning journey!
Team Treehouse Impressions
As a complete new person to an overwhelmingly intimidating field, I really loved the clean interface and design of Team Treehouse. The videos all seemed to be professional produced as opposed to well-produced home videos from other sites.
At first, even though the quizzes were a surprise, I was really happy to have them. The quizzes were also intimidating, but it helped my brain learn how to comprehend the lingo of this new realm. Also, the payoff of correctly answering a quiz question when you think you’re lost was a nice additional motivator.
Later on in the course, I quickly realized that different teachers teach different subtopics in each track, and some include different quiz methods and teaching methods.
Quizzed Out
Since I’m a stay at home / work at home mother, after a while, the quizzes were a really big disruptor in my learning process and time management. I don’t have the luxury of uninterrupted time or much quiet space. So, sometimes I want to be able to power through retaining as much information as I can during my small gaps in free time. During those moments, I was very frustrated to have a three minute or four minute video followed by multiple quizzes (sometimes five quiz questions / challenges back to back).
I should probably mention that I have years of analyzing my own personal learning process, because before I taught college I had to learn how to teach, and before I could learn how to teach, I had to learn how to learn.
I’m a visual learner. I also tend to have “bursts” of energy or “pockets” of moments where I am ready to dedicate hours to a task or subject. This trait of mine made me and Team Treehouse’s pacing very off-putting after a while
This could also be partly due to the fact that since I’m self-learning, I have my toes dipped in a lot of different areas. Some books I’ve read are ahead of the beginner tract, along with some of the projects I’ve accomplished, yet still a lot of it is new. Moreover, through my troubleshooting and projects, I’ve learned that those are my quizzes. Those stick in my brain and motivate me more.
Moving Forward
I don’t mean for this post to come off as if I am completely hating on Team Treehouse. I really do think that they’re a fantastic resource, and their videos have helped me a lot. As a new beginner, I’m happy that I chose this platform because the quizzes were strict and did help me persevere through some frustrating challenges (which I’m sure is the point, to mimic the life of a developer!).
However, for myself, I have decided to take a pause in my membership with Team Treehouse and move on over to try Udemy.
Udemy Impressions
First, I love the freedom of Udemy. I can skip forward in a course if I want and from what I can tell–there are no quizzes! I will say that Udemy is more attractive to me now because I am more experienced in this realm and have already personally decided to stick with this learning journey. However, if I was a new person and unsure, I think that Udemy’s platform could be potentially intimidating and perhaps give new students too much freedom to get in over their head.
What it really comes down to is personal preference and situation. For my personal situation, I’m excited to have the freedom to be able to utilize my time as best as possible, at my own speed. That will be a big plus for me!
I’ve been very interested in the topic of machine learning, so I did some research and found an awesome project that I could do with my Vector Robot (…who I may refer to as El Robo sometimes).
Set up your Google Vision account. Then follow the Quickstart to test the API.
Clone this project to local. It requires Python 3.6+.
Don’t forget to set Google Vision environment variable GOOGLE_APPLICATION_CREDENTIALS to the file path of the JSON file that contains your service account key. e.g. export GOOGLE_APPLICATION_CREDENTIALS="/Workspace/Vector-vision-62d48ad8da6e.json"
Make sure your computer and Vector in the same WiFi network. Then run python3 object_detection.py.
Step one was easy-peasy, since I had tested and ran the SDK a multitude of times previously already (woo-hoo!). But oh boy, guess what Step 2 truly looked like once you followed the Quickstart link for the API setup:
From the Service account list, select New service account.
In the Service account name field, enter a name.
From the Role list, select Project > Owner.Note: The Role field authorizes your service account to access resources. You can view and change this field later by using the GCP Console. If you are developing a production app, specify more granular permissions than Project > Owner. For more information, see granting roles to service accounts.
Click Create. A JSON file that contains your key downloads to your computer.
Set the environment variable GOOGLE_APPLICATION_CREDENTIALS to the file path of the JSON file that contains your service account key. This variable only applies to your current shell session, so if you open a new session, set the variable again.▸Example: Linux or macOS▸Example: Windows
Yeah, so things got complicated real fast. Instruction Set A:PT 2 is where I officially fell down the rabbit hole of Google Cloud.
For ease of reference, I’ve labeled the instructions sets A and B, but let’s be clear: All of B is actually just a sub-set of A. Instruction set B is just to set up Google Cloud Platform account and installation, which was not something I was anticipating to learn throughout this journey, but I am so happy I stumbled upon it!
Google Cloud is offering one year free for new users, which is a crazy value considering the retail price allotted is $900+ for the year!
Once I set up my Google Cloud Platform account and logged into the console, I felt a unique sense of adrenaline, almost with a tint of rebellion.
Reading all of the tabs and sub-tabs like: Cloud Build, Big Data, and last but not least… Artificial Intelligence. I don’t even have the right words that could capture the feelings I was experiencing while scrolling up and down the console’s navigation menu, but I definitely had a huge smile on my face. Although I had (and still don’t) exactly know what Google Cloud can do or how to use it fully, what was clear right away, even as a complete novice, was that this tool is powerful and if learned properly, it could be utilized for epic projects (and help with future career prospects, *wink wink*).
I‘m a very visual learner, so seeing the categorization of the navigation menu taught me vital information for my coding journey, because it can be used like a road map. Now I know for sure that two areas of interest I have going forward involve Big Data and Artificial Intelligence.
As I continued to read Google’s instructions, a little voice was starting to whisper that perhaps this project was over my head. Especially once I got down to Instruction Set B: Step 5
Set the environment variable GOOGLE_APPLICATION_CREDENTIALS to the file path of the JSON file that contains your service account key. This variable only applies to your current shell session, so if you open a new session, set the variable again.
Which corresponded to Instruction Set A: Step 4
Don’t forget to set Google Vision environment variable GOOGLE_APPLICATION_CREDENTIALS to the file path of the JSON file that contains your service account key. e.g. export GOOGLE_APPLICATION_CREDENTIALS="/Workspace/Vector-vision-62d48ad8da6e.json"
And these instructions left me like this:
But no matter how much I re-read the instructions, I just couldn’t understand what they were telling me, which actually made me feel as if I didn’t even know how to read! After some trial and error, I looked more like this:
Finally, I decided to cave and ask for help. Luckily, my husband and his coworkers were over. I asked them for help but quickly came to learn that Google Cloud is a pretty niche subject to just casually ask about.
“Damnit, Jason!”
I was stuck on the .json file part, for which I kept yelling “Damnit, Jason!” in the house, confusing everyone. (“Jason” has now become an inside joke in the house, and from now on, whenever I work with .json files again, this first memory will be in my mind).
The next day while cooking dinner, I decided to watch a YouTube instructional video on how to setup Google Vision. As the tutorial leader was going through the steps, I was happy to see that some of the steps were very familiar, which meant I had definitely learned something from the previous day. And then finally, I got my answers, and the instructions were demystified. I was so, so happy to finally understand what the heck the above steps meant and looked like.
What I love most about coding and these projects is that I end up learning about so much more than what I initially intend.
Through this process I learned about:
Google Cloud Platform
How to install and configure Google Cloud Vision API
Virtual Environments and Environment Variables
Changing .bash profile in the terminal (including how to inadvertently fuck up your bash profile, and then fix it again).
More about python pip installations and configurations
How to move through different directories in the terminal
Finally, I had made it passed every single step and it was finally time to test out the python program to see if my Vector Robot would begin object detecting.
Successfully setting up the Google Cloud Vision API was such a pivotal milestone for me in this learning journey.
Now, that means that I had just gotten past one subset of the main instructions for my intended purpose of setting all of this up, which was ultimately to run one program, created by one single person, and see if it worked.
Drumroll…anticipation…andddddd:
It did not work.
Type Error:__init__() got an unexpected keyword argument ‘enable_camera_feed’
I looked over all of my steps. Everything was in order, but I kept getting a Type Error from the python file written by the creator. After countless web searching and troubleshooting, I was at a dead end. I had only one option left. I had to contact the person who made the program.
I went to GitHub and posted my first “issue” on the python file. This was the first time I had ever posted an issue, so I was very unsure of any “etiquettes.” The creator got back to me and said that he had created and tested this back with the old Anki Vector SDK package, and not the most recently updated one.
This may seem like a very sad ending to this story, since the project was not successfully executed on my end, but I was so happy with that response from the creator! The reason being because the creator didn’t say that my error message was due to something wrong with my steps, but perhaps elements outside of my control.
What a relief! For that alone, I considered this project a “success” in that I followed through to the end.
Although I couldn’t get this program to work, that doesn’t mean I can’t try to write my own program utilizing similar features. This project taught me a lot of different lessons and I can’t wait to try more Google Cloud Platform projects while I have the free trial!
Kris Learns Coding 