Creating a Computer Graphics Career Goal: A Rough Plan for 960 Hours

[Ffflovin_20]

TL;DR Summary

Is it feasable redoing 6 courses in a 4 month period? --Split into 2 months each with 3 courses. 1st half: Linear Algebra, Calc1, and C intro. 2nd half: (Either OOP or Algorithms), Data Structrues, and Calc 2.

Goal/Background

I've taken these courses before(from different college/sources but didn't resonate)

So my current career goal is to get into computer graphics industry or research.

• Open to both game engines or simulation graphics

Academic goal is to prepare myself for upcoming Computer Graphics Course

I have the summer to review and brush up on fundamentals,

with appox. 960 hrs at most. (10 hours x 6 days ideally)

I'll now detail my "rough plan"

Rough Plan (For first 2 months)(approx 480 hours)

1.)Linear Algebra (https://ocw.mit.edu/courses/18-06sc-linear-algebra-fall-2011/pages/syllabus/)

• COurse says atleast 150 hours required. Though I might realistically be shorter due to my previous exposure

2.) Intro to C

• -Also includes some shell scirpting.

• -Course takes 14 weeks, although that's assuming a full courseload of 5 classses. So I can probably compress it to 7-8 weeks.

3.) Calc 1/ Differential Calc()

• -PRof leonard's videos are about 50 hours in total

• -I will use Spivak's Book for practice(crazy IK)

• -I already have some proof experience

• -if my proof are actually more "basura" than I thought, Ill just do the hard AF computational ones.

ROugh Plan (For last 2 months)(480 hours)

4.)HELP! Object Oriented() Or Intro to Algorithms(https://ocw.mit.edu/courses/6-006-introduction-to-algorithms-spring-2020/pages/syllabus/) ?

• WHich one will benefit me the most as pre-req for computer graphics?

5.)DAta Structure Course

• -Course takes 14 weeks, although that's assuming a full courseload of 5 classses. So I can probably compress it to 7-8 weeks.

• -Linked Lists, Trees, AVL trees, Heap trees, Radix Sort.

6.)Calc 2/ Integrations( )

• -NOt sure if Spivak covers this.

• -if not, I'll just resort to the "McDonald's" of calc books(ie. stewart, thomas, etc)

Closing

Ideally Multivariable Calculus should also be here, but theres a time constraint so...

Please tell me if this is a feasable strategy (Assuming average intelligence and an ability to prevent distractions)

• Also open to a more streamlined version to CG.

REason for these choices is that, I'm also going to take Probability course and OS, and also Proof based Linear algebra as well in the future.

I repeat that all of these I've already taken, but did not resonate/ stick in my head.

THanks for reading.

 

[berkeman]

Welcome to PF.

So my current career goal is to get into computer graphics industry or research.

Others will be able to help you with your questions, but can you say more about what kinds of computer graphics work or research you want to do? Are you talking about game engines, or FEA type graphics for simulation software engines, or realistic rendering engines (like Blender, etc.) or something else? That may help to guide some of the responses that you get. Thanks.

 

[Ffflovin_20]

Welcome to PF. Others will be able to help you with your questions, but can you say more about what kinds of computer graphics work or research you want to do? Are you talking about game engines, or FEA type graphics for simulation software engines, or realistic rendering engines (like Blender, etc.) or something else? That may help to guide some of the responses that you get. Thanks.

Thanks for the recommendations, I've now specified which area of CG.

 

[FactChecker]

Since you say that "it didn't stick", it's hard to tell. If you understood them, most of those classes have fundamental ideas that should stick, but the little tricks might not. Maybe you should look at your old textbooks. If it looks like you just need to resharpen your skills, then there might be a good chance it will work. On the other hand, if the subjects look like you forgot everything then I think your plan is too ambitious.

 

[Ffflovin_20]

Since you say that "it didn't stick", it's hard to tell. If you understood them, most of those classes have fundamental ideas that should stick, but the little tricks might not. Maybe you should look at your old textbooks. If it looks like you just need to resharpen your skills, then there might be a good chance it will work. On the other hand, if the subjects look like you forgot everything then I think your plan is too ambitious.

Sounds good I'll try to skim through them. I might reduce the courses to 4 instead. Remove Calc 2 and Data structures

 

[jedishrfu]

You need to see this talk

 

[jedishrfu]

I posted the TED talk because your detailed roadmap just seemed too strict to follow.

If I were trying to recover the math I learned as a Physics major years ago, I would use a site like MathIsPower4u.com to do it and not setup such a strict order of learning.

The key to learning math is doing problems. The math site I mentioned covers math from middle school up first year college as a collection of ten minute videos where a problem is presented and solved before your eyes.

You can stop the video at any point and attempt to solve it and then compare your solution to his. You can also skip over problems that you understand already and focus on the ones that are more interesting.

You have to think this math journey is like a trail hike, you can estimate the time it takes but realize you may run into roadblocks and difficulties that will slow you down. But you’re here for the enjoyment of the journey not stressing out about keeping a timetable of topics accomplished. Go with the flow.

 

[FactChecker]

It's not clear to me how important these subjects are to programming computer graphics. Maybe someone with knowledge of that can comment.

 

[Mark44]

It's not clear to me how important these subjects are to programming computer graphics. Maybe someone with knowledge of that can comment.

The six courses listed are Linear Algebra, Calc1, and C intro, OOP or Algorithms, Data Structures, and Calc 2.

IMO, the classes most relevant to the stated goal of graphics programming are linear algebra and C or C++. Being comfortable with linear algebra topics, especially vectors and matrices, is extremely important. With C or C++, going beyond the intro level would be preferable. Beyond these choices, I would probably rate data structures over algorithms. but I'm not sure whether OOP would be all that useful.

For the math part, I'm not aware of graphics programming using much in the way of calculus, particularly integration.

 

[Vanadium 50]

As I understand it, you took these courses before, and they didn't "stick" so your new plan is to try aagin, only faster.

While I can't say it for sure won't work, it does not seem to be moving in the right direction.

 

[shivajikobardan]

As I understand it, you took these courses before, and they didn't "stick" so your new plan is to try aagin, only faster.

While I can't say it for sure won't work, it does not seem to be moving in the right direction.

Why doesn't trying to learn look like moving in the right direction?

 

[Mark44]

Why doesn't trying to learn look like moving in the right direction?

The skepticism about "moving in the right direction" isn't so much about the learning part, but trying to do it all in 4 months.

 

[ohwilleke]

Is it feasable redoing 6 courses in a 4 month period? --Split into 2 months each with 3 courses. 1st half: Linear Algebra, Calc1, and C intro. 2nd half: (Either OOP or Algorithms), Data Structrues, and Calc 2.

I've taken these courses before(from different college/sources but didn't resonate) . . .

I have the summer to review and brush up on fundamentals,

with appox. 960 hrs at most. (10 hours x 6 days ideally)

First of all, you should not expect to get more than about 20 hours a week of review work done. Beyond that point you cease to be productive. Part of the process is getting away from it, sleeping on it, and coming back to it. So, you have more like a maximum of 320 hours of time to devote to it, and there will inevitably be minor hitches or times when you are under the weather for a day or two over the course of 4 months, so you really don't have more than about 300 hours you can count upon.

Second, the amount of time necessary for different subjects is very dependent upon knowing what you don't know or had trouble with the first time around, and placing that in the context of your goals with regard to what you need to know.

For computer graphics, from the mathematics side, at a minimum you need to be very fluid with linear algebra.

As far as calculus goes, not all parts of calculus are equally important. You need a firm command of the basic concepts and the notation including vector calculus notation.

But, the mechanics of integrating complicated functions from scratch is not normally something that someone in the computer graphics field would need to do. If you can't find an integral worked out in a table of integrals, you either don't need it or would outsource that work anyway. Tools like integration by parts are probably not something you are going to need to use. In your shoes, I would devote less of your available time to calculus, and more to linear algebra and programming.

You are certainly going to need to have a firm command of what you were supposed to learn in OOP, Algorithms, and Data Structures to do computer graphics. But again, what you should focus on is key. You will be tempted to review the parts you grasped well, and to procrastinate and marginalize the parts you struggled with the first time.

This gets to a deeper problem. Self-study can be a good way to learn things. I used self-study to teach myself high school geometry, high school pre-calculus, linear algebra, three semesters of calculus, a semester of discrete math, a semester of physics (thermodynamics, optics, and shallow intros to modern physics), plus two semesters of college level accounting. But self-study is particularly inefficient in areas where you have trouble grasping some concept, even with published study aids in addition to a text book.

Something that someone who understands it can explain clearly in 45 minutes (15-30 minutes if it is interactive and you are present and engaged and really listening and the tutor is a good one) can take days of angst to grok through self-study if you are trying to get it without that kind of assistance.

When I was doing self-study, I probably spent a third of my time overall trying to get my arms around the 5% of the material or so that left me stumped the first time around.

Realistically, you will be wildly inefficient to the point of not making the effort really worthwhile, unless you have a tutor (or study partner in the same subjects with different struggles than you have), to help you overcome issues that were rough spots or conceptual leaps for you the first time.

 

[shivajikobardan]

I suggest you share your syllabus and if allowed previous year assignments (unsolved) of CG. That's the only way one can tell what you should study. I've studied CG and I'll tell you you don't need anything you mentioned here to just understand CG. You don't need to be an expert in these subjects to understand CG. But if you really want to make games using CG, I'd focus on data structures for those 4 or so months. DSA alone is useful subject as well.

 

[jedishrfu]

Here's something you might consider. This Youtuber used ChatGPT to construct a self-directed learning plan using the Pareto and Feynman learning methods and then had ChatGPT flesh out the plan more fully with other prompts.



One thing, she failed to mention in the video is ChatGPT's propensity to give wrong information that you need to independently verify which will help you develop critical analysis skills.

 

[bryantcl]

I know I may be a little late to the party, but I'll throw in my $0.02 USD. I don't think this was mentioned but it may be helpful to you that over the course of learning, build a portfolio with your work. My understanding is that CG is a highly competitive field (basing this on what others have told me) so a portfolio might help set you apart.

Best of luck,

clb