Path I: Suitable for those with the least self-discipline
Let's start with the people with the worst self-discipline.
Such classmates tend to be hot for three minutes. I was accidentally stimulated to learn Python in order to pursue a career in data science. He will immediately go to the library or bookstore and pick up a book "X Days from Beginner to Mastering Python" and start nibbling. As a result, X days have not yet arrived, and the whole process from entry to giving up has been successfully completed .
If you fail to persevere, you must be responsible for it. But the biggest problem is that you overestimate your own self-discipline. Such students, I recommend you to go to the Coursera platform and learn a very good MOOC - Programming for Everybody step by step.
I recommend this course because the quality of the course is really good. First of all, the textbook is good. The source of this textbook has a story. First Allen B. Downey wrote an open book Think Python: How to Think like a Computer Scientist .
This book is rated on Amazon like this:
Charles Severance thought it was too well written and wanted to use it as a textbook. So with the consent of the author, I borrowed a lot from the content structure of this book and wrote a Python for Informatics .
When Charles wrote this book, he also opened the iBooks format, which included his own teaching videos for students to watch and learn directly.
Later, Charles extended this book into a MOOC. Shortly after its launch in 2015, senior engineers in Silicon Valley were eager to learn.
Charles knows the art of course iteration. He continued to add content, improve the curriculum system, developed a course into a special course (Signature Track), and upgraded the textbook to Python for Everybody: Exploring Data In Python 3 .
In the current global MOOC word-of-mouth list, this course of Charles has always been at the top.
This special course explains the simple syntax of Python in a simple way, and also uses some basic tasks of data science to drive you to write simple projects in Python language. This solid training process builds your confidence and sparks interest.
For students with low self-discipline, the following feature is more important: All work has a time limit .
Courses on Coursera, the weekly tasks are very clear. If the correct rate of practice questions cannot reach 80%, you will not pass the test. By the deadline, if you don't complete all the exercises and course items, you won't get the certificate.
The teacher leads you in the front, the teaching assistant urges you next to you, the platform reminds you with a timetable, and the classmates on the forum are pushing you with peer pressure... Want to be lazy? Want to fish for three days and dry the net for two days? It's hard.
Path II: Suitable for people with moderate self-discipline
If your self-discipline is above average, then your options are wide.
Here I recommend you another MOOC platform called Datacamp.
My first exposure to Datacamp was in early 2015. At that time, I took Duke University's statistics course "Statistical Inference" on Coursera, and the accompanying exercises were on Datacamp.
I was very impressed by this platform at the time, because the code runs in a cloud environment. Learners do not need to install any environment locally, a browser that supports HTML5 standard can bring you a complete learning experience.
This is a great way to get started for beginners. You must know that many people's enthusiasm for learning is buried by the pit of environment configuration and dependent software package installation.
Two years later, Datacamp has grown even stronger. You can open the Data Scientist with Python learning path on the homepage to see the 20 courses already offered.
These courses cover everything from the basics of Python, to data manipulation, to artificial intelligence and deep neural networks.
All courses are designed to be short and concise. Generally no more than 4 hours, you can complete the study of a subject. This way you learn effortlessly and get feedback (automatic grading of practice questions) and a sense of accomplishment (certificate) in a fairly short period of time.
The course of this platform is completely controlled by the learners themselves . So I summarize it as suitable for learners with a certain self-discipline. It can not only give you instant feedback, let you know your position and progress at all times, not lose your way, but also fully experience the fun of self-learning.
Datacamp courses are generally free for the first part. Learning can only be unlocked after purchasing the latter part. If you are confident in your learning ability and perseverance, you can buy a full time period (eg one year) course. During this period, you can take courses on all platforms, and you can earn a certificate after passing. There are already discounts for such a purchase plan, and there are significant discounts at certain times of the year, which are very cost-effective. It is recommended to add more attention to the shopping cart.
This is the deep learning framework Keras course certificate I got at Datacamp. It really only takes a few hours to complete it, and the sense of achievement is quite strong.
Path III: Suitable for people with strong self-discipline
The aforementioned courses are expensive. The average price per course on Coursera is around $49. For groups of students from developing countries, Coursera can provide financial aid. You can fill in the application form truthfully according to your needs to get funding.
For students with strong self-discipline, your choice can become very simple and straightforward - you can use the most respected textbooks and study by yourself.
The most respected teaching material, in fact, does not exist. As a Western proverb goes:
One man’s meat, is another man’s poison.
There is nothing in this world that everyone agrees with. But there are textbooks with very good reputation, such as this strangely named "Learn Python the Hard Way" (Learn Python the Hard Way).
Don't be fooled by the name, this is a poor introductory Python tutorial. On the contrary, the design of this book is very suitable for people's cognitive laws.
We learn things, from shallow to deep, from easy to difficult, and gradually progress. If you blindly pursue new knowledge, then what you have learned before will be quickly forgotten. If you are always spinning in place, it will bring a dull and boring feeling. Do you still remember the year paper you did in your third year of high school?
A good textbook should provide learners with new knowledge and content in each chapter and provide enough challenges. But the challenge should not be so high that the learner becomes frustrated and gives up. At the same time, it cannot be ignored that in the follow-up content, the knowledge previously learned will be changed and repeated in a continuous spiral. Only in this way can we consolidate what we have learned, let learners feel the role of basic knowledge, and enhance the pleasure of learning.
This is a bit abstract. In fact, there is an English textbook that is very consistent with the above cognitive laws. It is this set of textbooks that I have repeatedly recommended in the classroom and in the article:
Learning Python the Stupid Way is one such book. All you need to do is open the book, open a useful code editor at the same time, and start typing, running, and changing the code as required by the book...
The picture below is the code I typed according to this book when I was learning.
The completeness of Python basic content training in the book is unparalleled so far . By the way, this book is available in Chinese. So if your English is not good, don't worry at all.
A word of advice, English should really be studied hard. Not only does it broaden your horizons, it also increases the chances you may gain. Considering that the readers who read this part carefully are very self-disciplined people, I don't need to say more.
challenge
The three basic Python entry paths are finished. With a clear understanding of your own self-discipline, I believe you can find a way that suits you to gradually learn and master Python.
But after completing the reading and listening to the class, is it all over? of course not.
Many people make mistakes here. They think that getting a certificate, or finishing the textbook, even if they really master Python. Then put the language aside and go to the American TV series and novels. Trust me, you will forget . If you never forget things you haven't touched for a long time...go to the hospital and check it out. Most people's memory patterns look like this:
Without intervention, within a week, you can almost forget all the new knowledge you have learned. What if you don't want your hard-earned Python knowledge to be wasted so easily?
practice
You should practice.
To practice Python skills, it is not necessary to find a core technology department of a Fortune 500 company, and it takes N years to complete the "996" work. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming.
I really feel like I've got my hands on Python after finishing my first Github project.
The project is very simple, using Python as the glue language to connect a series of tools together. You can change the content written in Markdown into various formats with one click.
Formats include but are not limited to:
- PDF/LaTeX；
- Word；
- Bitcron documentation;
- MarkEditor manuscript;
- MWeb documents;
- Bear manuscript;
- TextBundle (can import MindNode, Ulysses, etc.);
- Reveal.js slideshow;
- Release version Markdown (one-click image to Qiniu image bed);
- Local version of Markdown (remote Markdown such as Jianshu synchronizes pictures to the local);
- Day One Diary.
Some of these functions are being released in the Github public project, and the address is here. Accordingly, I have also written an introduction.
This little project, I started doing it in 2014. To be honest, looking back at the code at that time now, it is horrible. But if you get to yourself, you can find all kinds of interesting problems in life, and then think about whether you can use Python programming to solve it. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming. You can find all kinds of interesting problems in life, and then think whether you can solve it with Python programming. Your code has this feel, proving you're making progress.
Don't expect to write perfect code as soon as you shoot, keep the word "iteration" in your heart at all times. That way you can tolerate your own clumsiness and keep improving. As the ancients said:
Diligent study is like a seedling in spring, it does not increase, but grows day by day.
In the process of doing this project, I have encountered a series of problems such as Chinese encoding, privacy information storage, file name space processing, absolute and relative paths, publishing process division, function decoupling, parameters attached to Web image addresses... and so on. .
By reviewing the logs recorded with the Git version control tool, and the version comparison function, you can clearly see when and how to solve these problems. And don't forget to tick the new little skills in your toolbox.
When the small problems are gradually overcome by you, you can truly feel the value of the skills you have learned and accumulate confidence bit by bit.
At the end, I will also give you a python spree [Jiajun Yang: 419693945] to help you learn better!
