Computer Log

Well so much for resuming the course away from my house. The video doesn’t show extra video options such as full screen or the video playback controls. In addition when I move my cursor over the screen, it dissappears. I tried switching the video driver to the open source one but it didn’t help. This remote computer is running Linux Mint 19.3 Mate. I installed Anaconda so I guess I can still practice.

For the previous post I talked about recreating my random customer SQLite database. I had rewritten this program in Julia and was happy with the speed improvements [see this post]. So because of my feeling of constantly fighting with Julia to work with SQLite, a short time ago I started work on getting it to run in Python 3. Which also included recently adding a random amount field.

Here is the last few seconds output, of my Python 3 program creating 1,000,000 records. As you can see it took about 30 minutes. So a substantial 45 minutes faster than the Python 2 program, and only about 10 minutes longer than the Julia program. To have a program that doesn’t break at every point release…I can live with slightly slower speeds!

998000 Written 2000 remaining
999000 Written 1000 remaining
1000000 Written 0 remaining
Finished!

real 30m16.161s
user 24m42.720s
sys 5m31.454s
(base) [bill@bill-ms7b79 CreateCustomer]$

This has been a great course…so far. It’s making me take longer than necessary because I keep stopping to try things on my own data. Which is a good thing. Also I can see where I did some of this back in 2018, including stuff with SQLite. This helps me solve my one issue with this course…no SQL. So I stopped again because a customer database I created doesn’t have an amount field which I want, so I can try out some pandas math methods. I have added a amount field to my random customer records but did not convert that to SQLite yet. I think having a lot of customers with an amount field will be fun to play with. It will allow me to select a various subsets of data, for example I can do amount calculations just on the people of Florida or Texans born in 1960 or Female Californians that are over 30 or a list of people who donated less than $1 or everyone just to see how python/pandas handles a lot of records…stuff like that there.

I defined the amount field as FLOAT in SQLite, which seems correct, I also saw REAL suggested. Viewing the amount looks like 28.35, 16.61, 7.18, 50.2 and so on. Summing a range of data produces the same answer in python/pandas as the LibreOffice Calc spreadsheet program.

Well as of today I must say the Udemy course, Data Analysis with Pandas and Python, which I only started yesterday, for me is better than Linux Academy’s Using Python for Data Management and Reporting. The author of the LA course wrote, I feel, a good intro article, Tricks for Working with Data in Python, which made me want to take a deeper dive into his LA course. However IMHO the quality of the course didn’t live up to his article. I spent considerable time documenting problems only to be left with the feeling that my words just ended up in the virtual circular filing cabinet. More time complaining than learning.

Positives of the LA course

It is run from the cloud, so you can resume your work anywhere.
It is focused on Linux.
I think it was also going to get into SQL.

However the above positives, especially the first point, are a general feature of LA and have nothing specifically to do with this author/course.

I must say, The original guy that I dealt with at LA [before they went A Cloud Guru], was very nice and accommodating.

Positives of the Udemy course.

The author, so far, explains things very good.
Cost. Very affordable! Cost given upfront.

Negative of the Udemy course

Since I spend most of my week away. I would have to reinstall everything on another computer.

I didn’t really mind installing the rather large Anaconda programs, because I’m getting very interested in data science. Also it’s fun playing with Jupyter Notebooks again…I feel I’m gaining a better understanding of it’s usefulness.

Signed up for and started Data Analysis with Pandas and Python at Udemy. I certainly can’t complain about the price, it’s less than $15. I watched the intro video and and it covers the things I’m interested in, Pandas, Anaconda and Jupyter Notebooks. Basically installing Anaconda installs everything needed for the course. So even though this material is offered at Linux Academy, of which I’m a member, with all the errors and their lack of interest responding to my issues with the course, I decided to try this course. I’ve looked at courses from them before but so far have never taken one. And for me a huge plus is telling me the price up front.

They also offer a “Certificate of Completion”. They do say Udemy is not an accredited institution, and as a result, the certificates cannot be used for formal accreditation. While that is important, to me what is just as important is that I feel that I’ve learned something. With my ability to generate many random fake but real looking data I can further explore pandas capabilities.

Well so far I’m very impressed with the instructor and the content. He covers installation on Mac and Windows only so I did have to take a small diversion from the course to install what was needed on Linux. But I’m guessing the usage should be the same.

I see many online computer courses that sound interesting. I go to check them out and I see something Like “enroll for free”, well thank you very much for collecting my info at no charge. But how about not wasting my time and first tell me how much the darn course will cost! By not telling me I assume it is expensive.

From Coursera website – Learn skills from top universities for free.
Then I read elsewhere “100 online courses from Coursera are $0 now through May 31″…so they’re not normally free.

Zoom reverses decision to only offer End-to-End Encryption for paid accounts.

Lots of errors in this (Using Python for Data Management and Reporting) course. I reported a small omission earlier. I think they use to say they’d get back to you…they don’t now. At the end of a lecture they ask if it met your expectation or needs improvement. I responded…needs improvement.

Here is what I reported…

The text below the video doesn’t match the video. In the video the major points seemed to be about loc, iloc, isin and the tilde(~). None of that was in the text below the video. Instead the text, repeated the text below the previous video, “Grouping and Counting Data” and “Slicing (Filtering) Data”. Also both this and the previous video includes “Creating a Basic DataFrame From JSON” which was covered several lectures previous in “Lecture: Creating a panda DataFrame and Examining Its Properties, Part 1”.

At this point the best thing I can say about this course is that it sparked an interest in me. At the very least it makes me want to look or further look into Jupyter Notebooks, Anaconda, pandas and dataframes.

It takes quite a bit of effort to write these complaints up. Especially a complaint like this where I had to go back an look at other lectures. Also, trying to double check things to make sure I don’t erroneously report something.

I had hoped to get the certificate of completion. But at this point I’m only 41% through with the course. I might learn more by playing with this stuff and watching YouTube videos myself.

I have to ask myself does The Cloud Guru buyout play any role in this?

Contacted LA about the above problem. However I did as the British comedy sitcom IT Crowd suggested and turned it off/on…reboot and restart, and it worked!

Saw this web article Tricks for Working with Data in Python, it looked interesting so I started the course Using Python for Data Management and Reporting. I always liked working with data so this seems like a good course to try. Topics covered Python, Jupyter Notebooks, Anaconda, pandas and dataframes, SQL, mongoDB, Working with Excel, LaTeX. These topics are also important in the data science field. So some things I’ve worked with…somethings I haven’t. Anaconda seems interesting. I briefly looked at Jupyter Notebooks while looking at Julia, but for some reason had problems seeing what all the fuss was about. So this course will help with that. Also looked at dataframes in Julia which I’m guessing took the idea from Python/Pandas.

I think a lot of people used Excel for some of this. And got locked into a inflexible proprietary solution. But in my short exploration of the Open Source solutions, Excel seems primitive by comparison. And I haven’t even mentioned Machine Learning! Still it may be interesting to work with Excel data.