At what point should I consider moving on from Python...

At what point should I consider moving on from Python? 7+ months of self-studying but I have found that Python is not nearly as respected as much as I thought within the professional world and I need a job. I have been focusing on data science and a lot of neat ML/DL stuff, but on the job front they need people with a lot more mathematical background in those fields.

I am interested in delving into some more 'formal' programming, like maybe C++. I'm interested in sound synthesis for music, but basically I just don't know where to go from here.

Attached: index.png (256x86, 2K)

C, C++, Haskell. In that order

go to sleep BW

I started with C for a bit and learned a lot of the basics there. Moved on to python because I was interested in data science.
Best course of action is probably jumping in to a personal project, right? Not sure if I should go for C or C++

>have been focusing on data science and a lot of neat ML/DL stuff
Where did you start with ML? All the resources explain shit like the readers are Rocket Scientists

I started with Andrew Ng's course. It's very maths heavy and I would only suggest starting there if you have plenty of time (I did). However the knowledge is fairly language-independent so you gain a good intuition for basic ML concepts.

The most straightforward, hands-on resource for me was Sentdex's machine learning series on youtube. The guy takes you through pretty much everything you need to know to work with ML in Python with the required mathematical caveats here and there.

I also use multiple books for reference (just basic Python for Machine Learning type stuff)

Avoid Siraj Rival on youtube. He's very enthusiastic, but will take any shortcut to get a video out there and I have never learned anything from him.

the only answer is to learn Java

You might not like this, but I'll break this down for you, because all the tutorials and shit online and on YouTube blatantly lie to you about the reality of getting a job in big data.

In real life industry, there are actually two "levels" of data science.
>actual data science
developing and prototyping ML models, exploratory statistics, etc. These people use R and Python and this is what most people think of, however this comprises like 2% of the industry and almost all of these guys have graduate degrees in math/stats/CS
>data engineering
designing, implementing, maintaining data pipelines from test to production. Believe it or not, these people predominantly use Scala and Java (due to the Hadoop ecosystem), and C++ traditionally although that is dying out. This is the bulk of the work and jobs in the real world

Everyone wants to do "data science" but you're basically fucked unless you 1) have an MS at a minimum in math/stats/CS or 2) have worked in a "data engineer" role for 3-5+ years. This is because prototyping basically takes no time in the long run of the project, and the skills that are a barrier to entry in that field are actually mathematics and statistics, not programming. I think it's also over represented by people like Andrew Ng because they're academics.

So my advice if you don't have an MS in stats or something, is to learn Scala, Java, and Hadoop shit, and study graduate-level mathematics and statistics on the side. If you want to go into the model-building kind of role, you'll lateral into it later, although you'll probably discover that "data science" is glorified database administration and you'll do something more interesting later on in life.

t. actual data shitter

The one on Coursera?

Not OP but I am the opposite,
I know high level math and have pretty good knowledge of machine learning algorithms, but I just don't know programming so I have no idea how to implement them
What would be the best option for me to learn?

what do you want to do? are you learning to make a living out of it?
each language has its perks and cons you use whatever is the standard for what you need, because it will make things a lot easier

>Python is not nearly as respected as much as I thought within the professional world
it's one of the most used programming languages
every language has a lot of hate you won't be able to find anything 'respected'

Hey man, I'm not here to disagree with you. In fact I've been learning most of this the hard way. (try going to a job fair and talking about data when all you have on your resume is Python, the looks on some people's faces is all you need to know you're going in the wrong career direction).

I would hvae nothing against doing data engineering, but as you said Python isn't so useful for that side. As much as I feel I have been 'duped', I've learned a shitton about programming in the process anyway and I'd like to do work where my lack of maths education doesn't hold me back so much.

Because I want a job I'm considering jumping in to Java as it seems every job posting and its mother asks for that.

Yeah, it uses Octave/MATLAB. There is a newer one with Python that looks a lot more updated. fast.ai also has a lot of tools for beginners.

python, sentdex tutorials

I want to make a living but I am enjoying what I am doing. Nothing about learning this stuff feels like a chore. However it doesn't matter what I do if I don't have a maths/CS degree when applying for Data jobs.

It's one of the most used, yeah, but if I'm not having any luck in data, and I have no interest in web dev is there a reason to stay with Python? Or should I not just move on to something like Java/C++ and work on some more in-depth software engineering?

fucking a

Consider your country age and education and think about taking CS or a 1-2 year programming course.
It's not about having only python in your resume is not having a CS background.
There's a lot of garbage in a CS degree but you also (in theory) learn how a computer/OS/compiler works, different methodologies and ways of thinking, working with other people, etc.
If you were hiring employees would you go with someone with 2-4 years of education or someone who learn by himself 7 months? Try applying to internships and if possible talk directly to the programmers to explain what you are capable of doing instead of HR reps, upload whatever you did to github, you worked on ML? Add some graphs with pretty colors to show off the results.

The frontend sucks, but working in the backend is pretty much the same anywhere, programming desktop applications, webdev and smartphone apps are all programming.
Java will probably cover most of the shit programming jobs where they accept newbies asap due to a shortage of people accepting their salaries, after 1-2 years of experience in one of those jobs companies will consider your experience even without a CS degree but they won't pay you as much.
I think the easiest way to get a job without education is showing off a good looking website but that's frontend.

I'm in the same boat as you, but with js. I learned the basics in c++, but wanted to make more "whole" solutions by myself. I think it's great as a hobby, but I would never want to work in webdev.
I guess your only choice with python is webdev (since you don't have a degree the science jobs might be a no go), which is too fucked up atm.
I'm going to learn Java. Feels like a good middle ground in abstraction, and has a good platform to be as general purpose as possible. Go into c/c++ if you want to do embedded programming. I don't know how good your foundation is, but syntax is the easy part. Implementation will always be key to learning something new.

I'm not op.
I have a me degree. I've done the math required in cs. Working in mechanical engineering sucks and the pay sucks, so I'm looking into switching into the software industry. Do you think it would be enough to have a strong portfolio, or is it wiser to go back to school for a while? I don't have any objections going back to school since I live in a country with free education.

join a startup that you think would benefit from ML and they'll let you implement it

[spoiler]and then ask themselves why their company failed[/spoiler]

I work in ML, and can give you a quick run down. You want to look at languages that can be deployed on a Hadoop cluster. Languages that work with Spark are currently the best.
>Python is #1 because everyone knows it and works on the cluster.
>Scala is a close second. Data engineers usually prefer this. It's type-safe and gives you more API options for Spark.
>R not bad, no one really writes it.
>Java will work, but at that point you might as well write Scala, which is pretty much the same thing.

Do you have a degree?

I mean 7 months isn't exactly a ton of time to be learning python "data" stuff. The numpy / pandas both have huge APIs do you know how to manipulate arrays / dataframes into the structure needed for certain processes in a data pipeline.

If I gave you a simple task like a putt-putt golf scoring sheet in a csv file could you get the counts each players unique score.

input:
player1,player2
2,2
2,3
2,4

output:

count
player1 2 3
player2 2 1
3 1
4 1

If you need a job go with Java. Pythons plenty respected professionally especially in webdev which will have a lower bar of entry then data science.

I think the problem is the industry meme-s datascience to hell and back when its really not as popular as people claim and requires knowledge of a lot more then python.

>Not sure if I should go for C or C++
What do you want to do with it?

Attached: depends.jpg (810x695, 100K)