/sci/ - Science & Math » Thread #10147568

160KiB, 925x550, meme22.jpg

View Same Google iqdb SauceNAO

Anonymous Sat 17 Nov 11:06:51 2018 No.10147568 View Reply Original Report

Quoted By: >>10147619 >>10147791 >>10147801 >>10147853

What do you all think of a Statistics vs Data Science undergrad major.

Also how programming heavy is Data Science? I fucking suck at programming (B- in a 1st year CompSci Java paper) and afraid of taking Data Science if it is a programming heavy major.

Anonymous

Anonymous Sat 17 Nov 2018 12:03:26 No.10147619 Report

Quoted By:

>>10147568
I've always wondered how some people can be good at math/science but then suck at programming. Programming is one of the easiest things out there. Fucking KIDS program in their pastime.

As for your question, university-level Data Science courses have a lot of programming exercises, but in the industry you don't have to program that much, because you can mainly use shit that others have made (libraries and statistics software etc.)

Anonymous

Anonymous Sat 17 Nov 2018 14:22:04 No.10147791 Report

Quoted By: >>10147888

>>10147568
In Data Science you should use Python and R instead of Java. Python is very easy language.

Anonymous

Anonymous Sat 17 Nov 2018 14:28:17 No.10147801 Report

Quoted By: >>10147812 >>10147816 >>10147817 >>10147826

>>10147568
How much programming should I learn?
I'm studying Economics as an undergraduate and our econometrics units have some basic R in them.
The university also offers some short courses in Python, would these be useful to take?
I'm unsure of the importance of programming, if someone has experience please share.

Anonymous

Anonymous Sat 17 Nov 2018 14:34:41 No.10147812 Report

Quoted By: >>10147816

>>10147801
Lol, how the fuck do you expect to analyze data without programming? It's all you're going to be doing.

Anonymous

Anonymous Sat 17 Nov 2018 14:35:43 No.10147814 Report

Quoted By:

The programming level in Data Science is simply being familiar how to use the latest packages and using fundamental programming knowledge such as iteration and conditioning. R and Python are the most popular languages due to being rich in packages and libraries.

You only need to know how to program when you want to create your own functions.

For example, you don't need to be familiar with Object Oriented Programming, algorithms or data structures.

Anonymous

Anonymous Sat 17 Nov 2018 14:37:03 No.10147816 Report

Quoted By: >>10147822 >>10149870

>>10147801
>>10147812

He could be using excel, STATA, Eviews, SAS etc.

Anonymous

Anonymous Sat 17 Nov 2018 14:37:26 No.10147817 Report

Quoted By:

>>10147801
Also, if you have a science degree and you can't code you're basically useless unless you did something experimental like geology, molecular cell biology etc.

Anonymous

View Same Google iqdb SauceNAO 1375445486647.gif, 2MiB, 295x254

Anonymous Sat 17 Nov 2018 14:39:08 No.10147822 Report

Quoted By: >>10147829

>>10147816
LMFAO
Good luck importing tensorflow into excel, fucking lol'd hard.

Anonymous

Anonymous Sat 17 Nov 2018 14:40:50 No.10147826 Report

Quoted By:

>>10147801
Most economists can't code for shit. I've met a lot of econ profs who are complete STATA monkeys and can't even use R. So it's not necessary.
But with that in mind, even just taking enough courses to get a minor in CS means you have substantially stronger computer skills than the average econ grad, and having the minimal programming skills necessary to duct-tape together a script in python is just an extraordinarily useful skill to have no matter what your job is.
Anyone in a desk job of any sort ought to learn at least basic programming.

Anonymous

Anonymous Sat 17 Nov 2018 14:44:44 No.10147829 Report

Quoted By: >>10147832

>>10147822
Ah yes because everyone needs to know machine learning.

Anonymous

Anonymous Sat 17 Nov 2018 14:45:49 No.10147832 Report

Quoted By: >>10147835

>>10147829
Ah I thought it was OP asking how much programming he should learn for data-science, my mistake.
Yeah if you're studying economics it doesn't matter, just make sure daddy's contacts get you a good job.

Anonymous

Anonymous Sat 17 Nov 2018 14:52:24 No.10147835 Report

Quoted By: >>10147881

>>10147832
https://medium.com/@thisismetis/dear-aspiring-data-scientists-just-skip-deep-learning-for-now-9c17030f6e44

Anonymous

View Same Google iqdb SauceNAO _90980653_2fb5401a-e706-4efe-881 (...).jpg, 25KiB, 660x371

Anonymous Sat 17 Nov 2018 15:17:53 No.10147853 Report

Quoted By:

>>10147568
tell the Big Data Nazis to take a Big Suck on my Penis

Anonymous

Anonymous Sat 17 Nov 2018 15:36:50 No.10147881 Report

Quoted By:

>>10147835
>skip deep learning for businesses, instead focus on all this other PROGRAMMING heavy stuff
fair point I guess

Anonymous

Anonymous Sat 17 Nov 2018 15:39:50 No.10147888 Report

Quoted By: >>10147906 >>10148030 >>10148140

>>10147791
R doesn't scale well for large data sets

Anonymous

Anonymous Sat 17 Nov 2018 15:49:38 No.10147906 Report

Quoted By:

>>10147888
Not even with tidyverse etc.?

Anonymous

Anonymous Sat 17 Nov 2018 16:45:05 No.10148030 Report

Quoted By:

>>10147888
R is designed for large datasets what are you talking about?

Anonymous

Anonymous Sat 17 Nov 2018 17:34:39 No.10148140 Report

Quoted By: >>10148361 >>10149876

>>10147888
suck my cRusty foReskin you fag
bet you use hadoop

Anonymous

Anonymous Sat 17 Nov 2018 19:16:45 No.10148361 Report

Quoted By:

>>10148140
>yo just install this faggotizeR package and you'll be fine

Anonymous

Anonymous Sun 18 Nov 2018 08:27:56 No.10149870 Report

Quoted By:

>>10147816
thats a fucking business analyst. all my data sci interviews ask ML and leetcode questions. expect Python, Tensorflow, Spark, Scipy, OpenCV. hell even got asked how to optimize kernels in CUDA once

Anonymous

Anonymous Sun 18 Nov 2018 08:30:03 No.10149876 Report

Quoted By:

>>10148140
python is easier to integrate with enterprise systems and other Software engineers. if its just one off EDA it doesnt matter what you use.

Capcode	All Only User Posts Only Moderator Posts Only Admin Posts Only Developer Posts
Show Posts	All Only With Images Only Without Images
Deleted Posts	All Only Deleted Posts Only Non-Deleted Posts
Ghost Posts	All Only Ghost Posts Only Non-Ghost Posts
Post Type	All Only Sticky Threads Only Opening Posts Only Reply Posts
Results	All Grouped By Threads
Order	Latest Posts First Oldest Posts First

Your latest searches