Data Science, Data Analytics, Python, Noob


#1

Hello LCTHW nation! Special thanks to @zedshaw for promoting education, tech-ed and community outreach. I hope you and your loved ones are now safe after Irma!

My question here will begin with a brief summary. I am new to coding and want to get a job in data science. I have used multiple resources and am measuring my progress. However, I am still new and unsure as to when I’ll be able to find a job that could pay me $45,000 or more (and for what?).

Is data analytics a nearer-distance, with regards to finding work with only an associates degree in Liberal Arts?

Currently, I have finished taking quizzes in two Python iOS apps, completed 70% of Codecademy’s “Learn Python”, am only at ex14.py of LPTHW (2.6.5), and have many other resources. For instance, I am a Salem, MA native and frequent General Assembly, a non-accredited tech-school in Boston and beyond. Tonight I took an intro to data science course, and eventually want to take the $16,000 course in Full-Time Data Science Immersive, though if I take that calculated risk I will have to make sure there is a career on the other end.

My current accomplishments are scarce. I just finished my first program for a Daily Caloric Intake Calculator, something that is going to be a part of a bigger project for learning the ketogenic diet. I practice LPTHW exercises with it, such as raw_input() and turning responses such as age (i.e. 26) into an integer: int(raw_input()), so I am aware that I have a long ways to go (will love to see how much I know after I complete the book, but as others have posted, it is difficult getting far fast with full-time work and responsibilities limiting daily practice time).

Realistically speaking @zedshaw, is it possible to land a job in the data industry within a year? Also, have you ever considered writing a book on data science? I’m only now beginning to learn about descendants of Python, such as Anaconda, but out of respect to your book’s instruction, I wait to dabble, and have only downloaded it because last week’s Python 101 course taught us how to. I mean, I follow your instructions, but I did buy a ten year old book. Python 3 will probably be the only one to use within five years, but until then I believe it’s still of huge importance for certain employers that you know things such as differences between P2 and P3, even if they don’t ask you in the interview. It might just be required.

Hopefully this wasn’t a lame first post! Albeit verbose. Apologies. Take care!


#2

So, first up, I prefer to be very honest with people looking for work in programming. If you need a job, then you must go find a job now and not bank on the idea that you’ll be a programmer making 90k/year. There is zero shame in working any job you can, and it’ll alleviate a lot of stress while you’re trying to change careers. The very worst thing to do is to quit your job, sell your house, or similar on the off chance you might get a job coding.

Next up, don’t do the General Assembly thing. You should look into classes taught at your local community college instead. Community College (CC) is taught by professional teachers, with accreditation, and an ability to give you a real degree you can take to anywhere. In addition to that a CC is normally mandated to keep costs low or even free so they can train people needed for the workforce and prepare people for 4+ year universities. The other thing is a CC will have classes that are going to be at night or online so you can continue to work while you study. And finally, a CC will pace the material appropriately so that you can have the time you need to actually learn the material rather than just crank and be incompetent.

I then recommend that you pick one single resource to go through and do just that until you’ve completed it and learned to code. I’ll obviously recommend my books because I think they work best for your situation, but you could do the others too. Just push through and fully complete them, maintaining a solid pace every day. Go ahead and finish the Codecademy material, then do all of my Python book, and then see where you stand. I also recommend that, if you want to be competent in coding, you should learn 4 programming languages. You could do Python, Ruby, JavaScript, and R. If you learned those 4 then just about any class you’d take at a CC is doable.

As for your employment ability in data science: Coding is only a small part of data science. You will need to learn some fairly advanced math and statistics. This is another reason I say hit a CC rather than short classes. You’ll want to take classes in Calculus and statistics at least. When you get to your Calc classes, look up the book “Calculus Made Easy” by Silvanus P. Thompson (updated by Gardner). Get that and the Schaum’s outlines for calculus to have problems to solve and you’ll get a good start in calculus that will help you though it. After that, I recommend people take statistics from the Sociology department if they want to go into data science. Data science is really just applied sociology to insane amounts of data. But, make sure you take a qualitative sociology class. If they have you doing ethnographies then you are in the wrong class.


#3

@zedshaw, what do you think about moocs?


#4

The moocs are getting better, but I see them catering to people looking for expert one-off topics and not really organized courses of study.


#5

First off, I would like to thank you,@zedshaw , for responding to my question. I love how helpful the tech community is. It’s really cool hearing feedback from someone whose work output I respect and admire.

I am going to finish LPTHW first, mostly because I hate unfinished projects. Data Science was a way for me figure out what I wanted to go back to school for. In truth, I had hoped there was some alternative to school, with respects to increasing work input in order to decrease time and money investment. However, I have documented your response and made notes of how I may likely alter my course.

See, I’ve already graduated from CC and could probably get financial aid for a 4-year school such as UMass in the computer science field. I currently work in retail as a computer salesperson (not at Best Buy, but I actually rep a 3rd party company who trains and sells Lenovo computers across the entire region). I got the job in March and knew jack about tech, so forcing myself to learn with like a deadline or good old-fashioned goals has helped my learning progress.

In truth I was bummed by your response, but only in the way that the truth hurts. I am a truth seeker and would rather be told by someone with tons of experience that the coding bootcamps offer a low likelihood of me finding work in data science. However, and this is significant, I had been told by my former statistics teacher that data analytics would probably be easier for me to get to, just due to the fact that DS is typically reserved for those with more advanced degrees, if not in mathematics or CS, then for those who have incredible researching abilities.

Regarding boot camps in general, would you say that they are high-risk, low-return for someone trying to get a job through one of those with limited initial experience? I mean… now that I know I should add 3 or more languages onto Python, I’m up for the challenge, but of course I want to focus on jobs I want that would benefit from them. Would you say that, scratching data science, even taking a boot-camp for data analytics would likely prove unrewarding, if my expectations were to get a job on the other end, and my initial knowledge was limited to, say, Python, Stats, a Liberal Arts degree, and a general understanding of what the field encompasses?

Thanks again for your response. I do believe that learning languages will become easier once I’ve learned the first one. I believe R and Python are the top two on the website https://www.kaggle.com/datasets , so R I feel would be my next goal. In truth, I am aware that I don’t want to become too expansive in coding (at least regarding skills acquisition for my next job, and for affording education), because I’d like to - quite ambitiously - jump from CS and mathematics to neuroscience. However, I’m as stressed as you can imagine when I’m not making the most of my time. I can rarely play video games even because I always feel like I could be doing something more useful. I just want to make sure I’m on the right path, but also the quicker path, to a career where I can afford to pay for school (or at least cover loans) while paying for an apartment, while building on that and so on.

Apologies for the ramble. I know I can just head to my local career center :sweat_smile:


#6

Next up, don’t do the General Assembly thing. You should look into classes taught at your local community college instead.

This advice is golden. Unfortunately, I can speak from personal experience. Sites like General Assembly etc (the “hardcore” “bootcamps” waving the high salary employment carrot) are nothing new. Private schools blossomed during the dot com boom/bubble/bust 20 years ago and did the exact same thing. My experience was likely not untypical - spent a lot of money on an “intense”, 5 month, full-time program, took away nothing of any value other than disappointment and exhaustion (eg the program spent an extra week on C. The first month was spent “covering” C and C++). Ironically, I was the only to gain employment in coding directly after, because I was the only one to show up for the interviews. But I was in no way qualified and that became apparent soon enough. The good choice for me would have been to invest in a 2 year college program. The private school I went to no longer exists.

I am back taking another run at this “coding thing” :slight_smile: but from a very different angle. I know I have much to prove to myself before I even think about a future in this. I was attracted to LCTHW because it required me to go in and start figuring things out - iow, it brought the coding and the coding environment to the forefront. Which seems to slow down the process and thus takes me deeper into an actual learning experience. In fact, OP, your comment about not having enough time struck me - I find that having large swaths of uninterrupted time is not necessarily better for learning. I am finding briefer, but consistent, bursts of time are working well for me. When I have too much time I tend to wander, or hack away mindlessly. :slight_smile: (or ramble on in forums :smile: )

But yes, to the OP: do not go with the bootcamps. And I would even suggest removing yourself from any mailing lists, because they will market to you, they will try to make it irresistible. They are a business and they exist by successfully selling education. /2c - didactics unintentional :slight_smile:

(EDIT: this was meant as reply to OP)


#7

You graduated from a CC but, not with a degree in Math, Computer Science, or similar right? That’s actually great news because it means you could go back and get a degree in CS but only spend 1 year doing it since you have most of the pre-requisites done already. But, if you can get a degree from UMass and have it paid for then that is definitely your best option. Before you go do that though, keep in mind that a CS degree is difficult because most CS departments are designed for kids who’d been coding since they were 12. They don’t need to be, but the education is all about “weeding out” potential students rather than teaching them in a way that determines their capability gradually.

But, I think what you’re looking for is a shortcut to get into the programming field, and for some reason you think Data Science is the shortest path? I’m generally curious why you’re focused on data science. It seems like the most difficult path since it involves many different difficult to learn things at once, so is there something about it that attracts you? And, try to be honest. Is it that it seems like something smart people do? Is it that, by being a very difficult goal, if you fail at it it’s not your fault? Did you read something that glamorized data science?

Let me know.


#8

Hi Zed! Sorry for the delay - I had to work.

You are correct, I graduated from a CC with a general education (Liberal Arts: Philosophy, Lit, Math, Sci, etc.). Reading this is great news! I am going to look into going back in the Spring, based off your recommendation. I appreciate the heads up for CS departments, but as long as I can take courses online I should have no issue.

To answer your second question, I don’t think Data Science is the shortest path. My story’s condensed version is below.

2017:
February: I work a job fair for a tech company and learn from a recruiter that a Bachelor’s in CS and 1-2 internships should make placing into a $60,000 a year job rather easy.

March: I lose my job and gain a better one. I train, merchandise, and sell Lenovo computers through third party, a job I began while knowing nil about computer specs.

May: After applying for a scholarship to UMass (and not getting it), I graduate, focus on work (after receiving full-time), and look into free workshops at General Assembly (search ga.co if curious).

January: Ok, I lied, I had been looking into General Assembly for some time. I took a free CSS/HTML/Java course there once and loved it. Learned easily (my biggest obstacle is money when it comes to learning, else I’d quit my job and go to school tomorrow). Looked into/was contacted by them about my interest in their Data Science programs. I took the Info-session on DS and it was up my ally. Random thoughts occurred in my head, such as using DS to prove gerrymandering occurred in specific locations, or revealing that cannabis regulation coincides with utilitarianism. Of course, these “social changes” were hobbies I’d concede. Nonetheless I decided finding a job in this field. So, I emailed them and learned that their basic recommendations for the full-time, $16,000, 9-5, Monday-Friday “Data Science Immersive” course was essentially a basic understanding of Python and statistics (i.e. histograms, and they’ll quiz you to see if you’ll pass into the course). One of their resources led me to Codecademy. Another led me to LPTHWay.

Fast forward to now, where I’ve been learning Python since then. My former Statistics teacher met up with me last month and showed me his work in R and gave me a bunch of great sources, such as Udemy, Coursera, etc. Resources aside, he also said that Data Analytics would probably be a little easier to get into, without some major degree.

My main issue is that I’d like to get a better paying job, not huge, I mean $45,000 is a lot but enough to pay for school and an apartment, but in general data interests me because that is where we are heading, the tech way. If everything is matter, then isn’t everything data by default? If we can understand data better, then can’t we understand everything better? Its potential interests me, and also that I hate retail, cooking feels wasteful when it’s full-time and I feel like I could be applying myself to greater goods, and smart isn’t the whole of it. I think some of your online videos show that, even though many smart people work in tech industries such as code and data sci, they are not always good at communicating that knowledge to other people (beginners or non-users). You excel at this, and I have a knack of transforming the complicated and complex to the simple-enough-to-understand levels for people as well, when I understand it well enough myself.

I am confident that I am more than smart enough to succeed in the data science field, and that the world needs more data science communicators (among other things), just as we need doctors and firefighters. I have weaknesses, but fortunately, taking responsibility is not one of them. If I failed it would be my fault, because it would mean that I quit. Failing happens all the time, every day, especially while learning Python for me. Oddly enough, frustration is a good sign I’ve learned, though it’s still weird having that silver-lining notion in the back of my head as I want to punch my monitor in its face.

I did not read anything glamorizing DS. In fact, I was frustrated at its explanations, similar to how you vent yours regarding what OOP is (by the way, I love your videos! Didn’t know what you were talking about, but I tried to match your enthusiasm and was able to follow along. “The Web Will Die When OOP Dies.” I don’t know much about it, but I’d say your prediction would be possible if two forms of the web were available at once. One that people were used to, and the other one you were talking about, until the old web was something people had to actively think about to remember.)

I digress.

I have definitely been asking around, while remaining flexible with my ambitions. The software engineer who instructed last week’s Python 101 and Intro to DS workshops told me “While there are definitely Data Science positions that require advanced degrees in Mathematics/Computer Science, they’re in the minority. Most DS jobs (and software jobs in general) simply care if you can do the work or not — there are lots of folks coding for a living with degrees in other fields or no degrees at all). One of the ways in which GA can help is by pointing you to companies that do skill-based hiring, the ones that want to see motivation and velocity more than CS degrees or years of experience. Just start asking around!”

In short, while CC seems to be the best option, by the sound of it, and most likely the next step, it’s the skill-based hiring that interests me the most. That’s going to be one of the next things I look into, but as to where yet I’m not completely sure. For me, I’m trying to get to $45,000 a year and am fighting off impatience. The only thing that matters to me is having realistic expectations, so even if it will take a couple years and CC is the next step, I will still be looking for opportunities throughout, adding onto and sharpening my skills all the while.

In closing, once again, thank you! Thank you via infinite loop.

I know you’re quite advanced in this all, but here is a website you may enjoy https://www.codenewbie.org/blogs. I love her podcast, and she is open to recommendations for guest requests. Would you be opposed to talking code on a podcast! If not, I’d try to give you a shout out on twitter. It’d be a great listen for me and thousands of others for sure.

Have a wonderful week Zed!

J


#9

Hello @zedshaw … it has been some time since I last checked your forum:

I have since bought dual monitors with a desktop, and am getting into your LPTHW 3 videos, so I am splended with my new learning environment.

I definitely branched out since last we spoke, but I wanted to see how intimidating and confusing data science would be, before returning to Python, so I’m glad it’s the challenge I thought it was.

Just posting this because your post made me ask myself whether or not I was doing this just for its difficulty and whatnot, so I spent some time soul searching, which led me to this one question for you:

Which voice recorder is that? I’d like to download it for when I become active on YouTube. I’m currently blogging, but would like to keep free knowledge flowing (not yours of course, I spent $60 on yours, but rather once I learn lists it helps for me to blog it and teach it right afterwards, not just use it)… I know there are tons out there, but I would like to use the type Zed A. Shaw uses!

Thanks


#10

Long time no see!

That is Camtasia 9, and quite honestly it’s crap, but all the screen recorders are crap. If Adobe Premier did screen recording I’d just use that. In fact, I may just use OBS to record the screen then edit in Premier. That might save me a ton of time to be honest.

Biggest problem with Camtasia is that it just forgets random things which end up costing you big. Worst is that you’ll be recording just fine with your external mic, and then one recording decides to switch to the internal mic for no reason. Or, you’ll do 100 recordings with an external mic and then do a voice over and it picks the internal mic. Or, you’ll save to a directory 100 times and then it decides that because you opened one project through finder you want it saved somewhere else.

Camtasia also botched their mp4 encoder so that it ruins all the colors, which is very important when you’re trying to do art, design, or similar videos. Even in mine with only code it just really makes the editor and intro screen look like total garbage. My solutions is to simply encode to .mov which gets the color perfect, then I use a Perl script that runs Handbrake CLI to do a way better rendering to mp4 with better compression.

Screenflow isn’t any better, as I found it was missing important features like simple video editing transitions and basic editing controls.