by
You’ve just started your first linear algebra course. The syllabus names one textbook and your seniors swear by another. A search for the best linear algebra books throws up dozens more, each promising to make vectors and matrices easy.
Starting with the wrong one makes the first few weeks harder than they need to be. A proof-heavy book turns simple matrix work into abstraction, and a dry one turns it into steps you memorize without understanding. Either way, it is easy to decide early that you’re bad at math.
My advice is simple. If you’re new to linear algebra, I’d start with one book: Gilbert Strang’s Introduction to Linear Algebra. It teaches the computation and the picture behind it together, and Strang’s free MIT lectures follow the same chapters.
There are 2 exceptions. If your course already names a textbook, I’d start with that one instead. If you’re learning linear algebra for engineering or data work, Boyd and Vandenberghe’s Introduction to Applied Linear Algebra is the better first book.
Linear algebra was a core part of my B.Sc. and M.Sc. in mathematics, and I later taught algebra, including matrices and determinants, to NDA and competitive-exam students. I haven’t worked through all 10 of these books cover to cover. For the newer applied titles, my verdict comes from their contents, their exercises and the course each one is built for.
Quick Comparisons
My suggestion is to find your goal in the table and jump straight to that book. The last column is worth a look too, because 3 of these books are legally free from their authors.
| Book | Best for | You should know | Free legal copy |
|---|---|---|---|
| Introduction to Linear Algebra | A first university course | High-school algebra; calculus helps | No, but the MIT lectures are free |
| Introduction to Applied Linear Algebra | Engineering and data work | Basic programming helps | Yes |
| Linear Algebra Done Right | A second, proof-based course | Proof writing | Yes |
| Linear Algebra (Hefferon) | Self-study with an answer key | College algebra | Yes |
| Linear Algebra: Theory, Intuition, Code | Learning through Python | Basic Python | No |
| Linear Algebra and Its Applications | A standard course textbook | College algebra | No |
| Schaum’s Outline of Linear Algebra | Extra solved problems | A first course in progress | No |
| No Bullshit Guide to Linear Algebra | Adults restarting math | High-school algebra | No |
| Linear Algebra for Data Science, ML, and Signal Processing | Machine learning and signal processing | A first course plus calculus | No |
| The Manga Guide to Linear Algebra | Getting past the fear of notation | High-school algebra | No |
Best Linear Algebra Books by Goal
Each heading names the job that book does best, so you can jump straight to yours. Every entry also says who I’d send elsewhere.
Best first course: Introduction to Linear Algebra
Gilbert Strang’s Introduction to Linear Algebra is the book I’d hand to anyone starting a serious first course. Strang keeps computation and meaning together: you row-reduce a matrix, and you watch what the column space and null space are doing while you do it.
His picture of the four fundamental subspaces is the most useful idea in the subject. Once it clicks, a big share of standard exam questions stop looking like separate tricks and start looking like the same question asked 4 ways.
The 6th edition from 2023 is current. Algebra fluency matters more than calculus here, and the book works best next to Strang’s free MIT 18.06 lectures. Strang retired from MIT in 2023, but those recorded lectures are still the best free companion any linear algebra textbook has.
It is my pick if you’re starting a first university course and will actually do the problems. I’d skip it if you want a proofs-first course, because that is Axler’s job further down. Read passively, it turns into a coffee-table book about matrices.
Introduction to Linear Algebra
- By Gilbert Strang
- Best for: a first university course
- Level: Calculus helpful
- Main strength: Balances computation, geometry, applications, and serious conceptual development.
- Watch-out: The book is strongest with Strang's lectures and a steady problem schedule.
- Verify the exact edition, format, and seller
Best for engineering and data: Introduction to Applied Linear Algebra
Introduction to Applied Linear Algebra by Stephen Boyd and Lieven Vandenberghe is the book I’d pick when linear algebra is a tool for engineering or data work. It builds vectors, matrices and least squares around problems such as:
- data fitting and regression
- classification
- control, tracking and portfolio problems
Theory arrives when a problem needs it. The book almost entirely skips eigenvalues, which tells you exactly who it is for, and a theorem-first math course needs a different text.
The complete PDF is legally free on Boyd's Stanford page, along with Julia and Python companions that turn the chapters into code. I’d read the free PDF first and buy print only if you want it on the desk.
I’d skip it if you’re in a pure mathematics course. The exam there will ask about eigenvalues and proofs that this book never builds.
Introduction to Applied Linear Algebra
- By Stephen Boyd and Lieven Vandenberghe
- Best for: engineering and data work
- Level: Basic programming helpful
- Main strength: Builds vectors, matrices, and least squares around modern applications.
- Watch-out: It is less suited to a traditional theorem-first pure mathematics course.
- Verify the exact edition, format, and seller
Best for proofs: Linear Algebra Done Right
Linear Algebra Done Right by Sheldon Axler is my pick for the second pass. It starts from vector spaces and keeps determinants out of the way until the end, so you learn linear operators as objects instead of as grids of numbers.
For a student who already computes confidently, it is the cleanest proof-based treatment in print. For a first-time student, it is the fastest way to decide the subject is not for them. The same design choice that makes it elegant as a second book makes it harsh as a first one.
The 4th edition from 2024 is Open Access, so the full PDF is legally free through Axler's site. Buying paper is purely a preference.
I’d skip it if you have never written a proof. My recommendation is to start with Strang for your first course and then come back to it.
- By Sheldon Axler
- Best for: a second proof-based course
- Level: Proof experience required
- Main strength: Develops vector spaces and linear operators with unusually clean structure.
- Watch-out: It delays determinants and expects mathematical maturity.
- Verify the exact edition, format, and seller
Best free textbook: Linear Algebra by Jim Hefferon
Jim Hefferon’s Linear Algebra is the best free textbook on this list and my first suggestion for anyone studying without a teacher. It is a complete, rigorous first course, and the author publishes worked answers to every exercise alongside it.
Those answers matter more than they look. A self-learner with no answer key can spend weeks practicing a mistake, and Hefferon takes that risk away.
The print copy has the same text as the free download on Hefferon's site. I’d buy it only if you study better on paper. The author won’t mind either way.
What you actually need is college algebra and patience. It moves more slowly than Strang, and for a beginner working alone that pace is a feature.
- By Jim Hefferon
- Best for: independent study
- Level: College algebra and persistence
- Main strength: Provides a rigorous full course with exercises and legal free access from the author.
- Watch-out: The print listing is optional because the same text is freely available.
- Verify the exact edition, format, and seller
Best for Python learners: Linear Algebra: Theory, Intuition, Code
Linear Algebra: Theory, Intuition, Code by Mike X Cohen is for people whose first language is Python. Concepts land through NumPy experiments you can run and change, which suits learners who trust a result more after computing it.
Basic Python is enough to start. It is good for checking intuition. The theory and the proofs still have to come from a main text, so I’d treat it as a companion to your main course text.
My advice is to work a problem by hand first and then confirm it in code. When you do it the other way round, the code does the thinking for you.
Linear Algebra: Theory, Intuition, Code
- By Mike X Cohen
- Best for: Python-oriented learners
- Level: Basic Python helpful
- Main strength: Connects theory with geometric intuition and executable examples.
- Watch-out: It is a companion for applied learning, not a substitute for a full proof course.
- Verify the exact edition, format, and seller
Best classroom textbook: Linear Algebra and Its Applications
Linear Algebra and Its Applications by David Lay, Steven Lay and Judi McDonald is the standard classroom textbook, and that is a compliment. It moves through matrix methods, vector spaces and eigenvalues in the order most courses test them, with large problem sets.
It has less conceptual daring than Strang and less rigor than Axler, with more predictable scaffolding than either. When a syllabus lists “Lay”, this is why.
The current edition is the 6th, and Pearson sells it with MyLab Math access. Before you buy, it is wise to check whether your course grades homework in MyLab. If it doesn’t, an older edition covers the same core for a fraction of the new price.
It is my pick if your course uses it or you learn best from worked examples. I’d skip it for self-study, where Hefferon gives you a comparable path for free.
Linear Algebra and Its Applications
- By David Lay, Steven Lay, and Judi McDonald
- Best for: students who learn through examples and standard course problems
- Level: College algebra and precalculus
- Main strength: Builds matrix methods, vector spaces, eigenvalues, and applications through a conventional classroom progression.
- Watch-out: The presentation is less conceptual than Strang and less proof-focused than Axler.
- Verify the exact edition, format, and seller
Best for extra practice: Schaum’s Outline of Linear Algebra
Schaum’s Outline of Linear Algebra by Seymour Lipschutz and Marc Lipson is the problem gym. Each topic gets a short review and then hundreds of fully solved exercises, at a fraction of textbook prices.
Speed on standard problems comes from doing a lot of standard problems. That is what I’d tell any exam student, and Schaum’s is the cheapest way I know to get that volume.
It assumes a main text. My advice is to use it next to your main textbook, never instead of it, because the concepts have to come from somewhere.
If you’re preparing for IIT JAM, CSIR NET or GATE, this is the book I’d keep on the desk next to your course text. Those exams reward fast, clean work on familiar problem types more than elegance.
Schaum's Outline of Linear Algebra
- By Seymour Lipschutz and Marc Lipson
- Best for: students who need many solved exercises
- Level: First linear algebra course
- Main strength: Offers concise review followed by a large bank of worked and practice problems.
- Watch-out: It works best as a supplement, not as the only source for concepts and proofs.
- Verify the exact edition, format, and seller
Best for self-learners: No Bullshit Guide to Linear Algebra
No Bullshit Guide to Linear Algebra by Ivan Savov earns its title. It walks from high-school algebra to university linear algebra in one compact volume, in a direct, informal voice that connects procedures to geometry and applications.
For adults returning to math after years away, that tone is often the difference between finishing and quitting. The compression that makes it readable also means it can’t replace a full proof-based text.
It is my pick if you’re teaching yourself and want straight talk plus computation. I’d skip it if your course already names a textbook.
No Bullshit Guide to Linear Algebra
- By Ivan Savov
- Best for: self-learners who want direct explanations and computation
- Level: Algebra and basic geometry
- Main strength: Uses an informal voice to connect algebraic procedures with geometry and applications.
- Watch-out: Its compressed style cannot replace a full proof-based university text.
- Verify the exact edition, format, and seller
Best for machine learning: Linear Algebra for Data Science, Machine Learning, and Signal Processing
Linear Algebra for Data Science, Machine Learning, and Signal Processing by Jeffrey A. Fessler and Raj Rao Nadakuditi is the advanced bridge. It grew out of their University of Michigan teaching. It goes from matrix theory to the SVD, low-rank approximation, optimization and estimation as data and signal work actually uses them.
A completed first course in linear algebra and calculus are firm prerequisites. A machine learning beginner who arrives without them will bounce off, and that is the book working as designed.
I’d save it for after Strang or Boyd. If you want the machine learning side without this much mathematics, my machine learning books guide has gentler options.
Linear Algebra for Data Science, Machine Learning, and Signal Processing
- By Jeffrey A. Fessler and Raj Rao Nadakuditi
- Best for: advanced students connecting matrices to modern applications
- Level: Calculus and a first linear algebra course
- Main strength: Develops decompositions, optimization, estimation, and signal-processing uses with current technical context.
- Watch-out: It expects stronger mathematics than an introductory machine-learning learner may have.
- Verify the exact edition, format, and seller
Best gentle start: The Manga Guide to Linear Algebra
The Manga Guide to Linear Algebra by Shin Takahashi and Iroha Inoue is the gentlest start on this list. It teaches vectors, transformations, determinants and eigenvalues through a story and pictures, and it works because it takes the reader’s fear of notation seriously.
High-school algebra is enough. The limited depth makes it a warm-up, so I’d pair it with a full textbook from this list as soon as the fear is gone.
The Manga Guide to Linear Algebra
- By Shin Takahashi and Iroha Inoue
- Best for: visual learners intimidated by matrix notation
- Level: High-school algebra
- Main strength: Uses a narrative and diagrams to make vectors, transformations, determinants, and eigenvalues approachable.
- Watch-out: The story format and limited depth make it a companion rather than a course textbook.
- Verify the exact edition, format, and seller
Free Linear Algebra Books and Courses
You can learn the whole subject without buying anything. If I were starting with no budget, I’d pick one free book: Jim Hefferon’s Linear Algebra, because every exercise comes with a worked answer. The rest of this list are alternatives for a specific course or goal.
- Jim Hefferon's Linear Algebra: a complete open textbook with exercises and worked answers.
- Linear Algebra Done Right, 4th edition: Axler’s current edition is Open Access, with the full PDF free.
- Introduction to Applied Linear Algebra (VMLS): the complete PDF, slides and Julia and Python companions from Stanford.
- Interactive Linear Algebra: Dan Margalit and Joseph Rabinoff’s free Georgia Tech textbook, with figures you can move to see what a matrix does.
- MIT OpenCourseWare 18.06 Linear Algebra: Strang’s lectures, assignments and exams, all free.
- 3Blue1Brown's Essence of Linear Algebra: the best visual intuition for transformations, determinants and eigenvectors.
3 of my 10 picks are in this list because their authors publish them openly. What a paid copy usually buys is print comfort, editing and a fixed sequence. The mathematics is the same.
How to Choose a Linear Algebra Book
I’d start from what you need the subject for next and then check whether you’ll have answers to test yourself against. Those 2 questions settle the choice faster than any bestseller rank.
Linear algebra is really two courses sharing one name. The right book is the one written for the course you’re taking now.
Do You Need Computation or Proofs?
One course is about solving systems, factoring matrices and computing eigenvalues. The other is about vector spaces, linear maps and proving why those computations work. Engineers and data scientists mostly need the first, and math majors eventually need both.
The question worth settling is which one the next 6 months require. A book built for the other course will feel wrong in every chapter, and that feeling has nothing to do with your ability.
Will You Have Answers to Check?
You learn this subject through problems, so I’d judge a book by its exercises before its explanations. A course student has a teaching assistant to catch errors. A self-learner with an answerless textbook has no way to know whether 3 weeks of work built skill or reinforced mistakes.
If you’re studying alone, I’d treat answers as a hard requirement:
- Hefferon publishes worked answers to every exercise.
- Schaum’s is solved problems by design.
- Strang pairs with years of MIT 18.06 exams and solutions.
Is the Book Set by Your Course?
If your course names a book, I recommend buying that book, including the edition. Problem numbers shift between editions, and a mismatched copy turns every assignment into a search.
In India, many B.Sc. and M.Sc. syllabi still list Hoffman and Kunze or Krishnamurthy, Mainra and Arora. If your university exam follows that syllabus, my suggestion is to read the listed book for its notation and keep Strang beside it to actually understand the ideas.
Books I’d Skip as a First Book
Some famous linear algebra books are excellent and still the wrong place to start. These are the ones I’d keep for later:
- Hoffman and Kunze, Linear Algebra: rigorous, dense and the M.Sc. standard in many universities. I’d use it as a reference in a second course.
- Halmos, Finite-Dimensional Vector Spaces: short and elegant, and written for readers who already think in proofs.
- Axler as a first book: the same book I recommend for the second course, for exactly that reason.
- Strang’s Linear Algebra and Learning from Data: a strong machine learning follow-up that assumes the first course is done.
- A new edition at full price: if your course has no online homework, an older Strang or Lay teaches the same core for much less.
How I’d Study Linear Algebra
One main book plus one free course is enough, and it is wise to finish that book before opening the next. Reading 3 introductions at once feels productive, but it delays the part that actually teaches you, which is solving problems.
The habit I’d build from day one is asking what a matrix does to space every time you multiply by it. Row reduction learned as a recipe works until eigenvalues, and then it falls apart.
For a First University Course
This is the order I’d follow:
- The first few 3Blue1Brown videos, so the pictures come before the symbols.
- Strang, one section at a time, with the matching 18.06 lecture.
- Every assigned problem by hand, before you look at any answer.
- Schaum’s, only if your exams need more practice than Strang’s problem sets give you.
- A weekly redo of the problems you got wrong, in place of reading new pages.
For Engineering, Data and Machine Learning
For applied work, I recommend this order:
- Boyd and Vandenberghe’s free VMLS, with its Python companion worked alongside each chapter.
- The 3Blue1Brown videos, whenever a chapter feels like symbols with no picture.
- Cohen’s Linear Algebra: Theory, Intuition, Code in place of VMLS, if you’d rather learn through code from the start.
- Fessler and Nadakuditi, only after a full first course.
For a Math Major’s Second Pass
For the proof-based second pass, this is the order that makes sense to me:
- A computational first course with Strang, if you haven’t had one.
- Axler’s free 4th edition for the proof-based view.
- Friedberg, Insel and Spence’s Linear Algebra in place of Axler, if your course treats determinants and canonical forms the traditional way.
Linear algebra sits next to other courses you’ll likely meet. If calculus is still shaky, I’d fix that first with the best calculus books. For groups and rings, my abstract algebra book picks are the next step, and my real analysis books cover the proof course that usually runs alongside the second pass.
FAQs on Linear Algebra Books
What are the best linear algebra books?
My top pick is Gilbert Strang’s Introduction to Linear Algebra for a first university course. Introduction to Applied Linear Algebra by Boyd and Vandenberghe is better for engineering and data work, and Linear Algebra Done Right by Axler is the best second, proof-based course. Jim Hefferon’s Linear Algebra is the best free option.
Which linear algebra book is best for beginners?
I’d start with Gilbert Strang’s Introduction to Linear Algebra, alongside his free MIT 18.06 lectures. If you’re studying alone without a course, Hefferon’s free textbook is the alternative, because every exercise has a worked answer. If notation itself scares you, I recommend a week with The Manga Guide to Linear Algebra first.
Which linear algebra book is best for machine learning?
I’d start with Boyd and Vandenberghe’s Introduction to Applied Linear Algebra, which is free as a PDF and has a Python companion. If you’d rather learn through code from day one, Mike X Cohen’s Linear Algebra: Theory, Intuition, Code is the alternative. Fessler and Nadakuditi comes later, after a full first course.
Is Linear Algebra Done Right good for beginners?
No. It is written for a second course and assumes you can already read and write proofs. If you’re a beginner, the better route is a computational first course with Strang, followed by Axler, whose 4th edition is legally free.
Are there good free linear algebra books?
Yes. Jim Hefferon’s Linear Algebra and the Open Access 4th edition of Linear Algebra Done Right are both legally free. So are Boyd and Vandenberghe’s Introduction to Applied Linear Algebra and Margalit and Rabinoff’s Interactive Linear Algebra. MIT OpenCourseWare 18.06 adds free lectures, assignments and exams.
Is an older edition good enough?
For self-study, yes. The core of linear algebra hasn’t changed, so an older Strang or Lay costs far less and teaches the same mathematics. I’d pay for the exact edition only when your course assigns numbered problems from it or grades homework through an access code.
Can I learn linear algebra on my own?
Yes, if you have answers to check your work. I’d use Jim Hefferon’s free Linear Algebra, because every exercise has a worked answer. When a section doesn’t click, the matching MIT 18.06 lecture usually fixes it.
Final Remarks
If your course starts this week, I’d get Strang and do the first section’s problems before your first class. If the pace feels too fast, Hefferon’s free textbook is the better switch, because it moves more slowly and answers every exercise.
Either way, my advice is to buy one book and finish it. A shelf of linear algebra books has never taught anyone an eigenvalue.
Best of luck.
Tell Google you want more of this.
Add Gaurav Tiwari as a preferred sourceOne tap, and this site shows up more often in your own Top Stories, AI Overviews and AI Mode. Remove it any time.
Disclaimer: This site is reader-supported. If you buy through some links, I may earn a small commission at no extra cost to you. I only recommend tools I trust and would use myself. Your support helps keep gauravtiwari.org free and focused on real-world advice. Thanks. - Gaurav Tiwari