This time, I passed the "Python3 Engineer Certification Data Analysis Test", so I will describe how to learn it.
・ Infrastructure engineer. ・ AWS qualifications have been acquired by professionals. ・ "G test" has passed (taken on July 4, 2020).
It all started when I tried to get an AWS expertise qualification earlier this year.
I thought that AI-related things that might affect business in the future would be good, so I aimed for "AWS Certified Machine Learning". However, even though I had knowledge of AWS, I had no knowledge of AI or machine learning. Therefore, I first obtained the G test, but I learned the importance of "Python" in machine learning and obtained the "Python3 engineer certification basic test". I took the "Python3 Engineer Certification Data Analysis Exam" this time.
The contents of the test are as follows.
Examination fee: General: 10,000 yen (excluding tax)
|Teaching material name||Classification||information|
|Python 3 Engineer Certified Data Analysis Exam(official)||Web(free)||まずはofficialサイトで試験内容を確認しました。この試験の特徴はテスト範囲が「Pythonによるあたらしいデータ分析の教科書」（翔泳社）から出題されることが明確になっている点です。どの章から何問出題されるか、まで明確になっています。例えば5章からは1問も出題されません。
The study methods and teaching materials used are as follows.
|Teaching material name||Classification||How to study|
A new data analysis textbook with Python(official)
Since it is described as a "textbook", it may be misunderstood if a beginner reads it, but in reality it is not described for beginners at all. At the beginning of the book, there is a statement that "the textbook presents the necessary information necessary for data analysis and explains it concisely", but the explanation is too concise and even beginners can not understand it due to lack of explanation. Because.
In order to understand the content, it is necessary to search the keywords in the book online or supplement them with other teaching materials. I feel that it is more of a "book for clarifying the test range" than a "book for studying".
|[Learn in one day] Python data analysis for engineers||Web(Paid)||This is a Udemy course. Usually 10,It costs 800 yen, but 90 during the time sale period, which is held about once every three months.%Close discount 1,It can be purchased for about 400 yen. I myself bought it at a time sale.
The video touched on installing Anaconda, using Spyder, Jupyter notobook, and basic Python operations, and was very easy to understand. The chapters before Numpy are also recommended for those who take the "Python3 Engineer Certification Basic Exam". However, in the later chapters, "Matplotlib" and "scipy" will suddenly become difficult, so I think it's okay to take a look at those areas and exercises at first.
Considering the test range of the data analysis test, "Foundation of Mathematics" and "scikit"-There is a lack of "learn", which needs to be supplemented in some other way.
After all the ease of understanding the video is the best.
|Learn at the prep school||Web(free)||「数学の基礎」の部分はWebで調べるより動画で勉強した方が理解しやすいことを「G検定」の学習の中で学びました。「Learn at the prep school」はその中でも特におすすめの動画になります。もちろん他の講師の方の動画もYoutubeに多数アップされているので、自分が理解しやすい動画を探すことが大事です。
|■ PRIME STUDY practice test||Web(free)||This collection of questions is the key to passing this exam. The official text is so wide-ranging that I have 100 about the content and keywords that appear in these three practice exams.%Emphasis was placed on understanding. Although it is stated in the official textbook as my study method, I will not be able to solve the problem that the keyword does not appear in the mock exam, but I divided it and bet on the mock question collection.
I thoroughly understood the contents of the mock question collection in the first time, and then repeated it. The last week of study was to solve this mock exam once a day. The following is the execution result.
|Practice test name||Score||Pass / Fail||Implementation date||result|
|1st practice test||047.5 points||failure||August 30, 2020||After one lap of the textbook. The situation that I do not understand at all. Learn to fully understand all problems.|
|1st practice test||100.0 points||Pass||September 06, 2020||Mock exam content After reviewing, retake the exam.|
|2nd practice test||080.0 points||Pass||September 06, 2020||なんとか初回でPass点超える。全部の問題を完全に理解できるように学習する。|
|2nd practice test||100.0 points||Pass||September 06, 2020||Mock exam content After reviewing, retake the exam.|
|3rd practice test||085.0 points||Pass||September 06, 2020||初回でPass点超える。全部の問題を完全に理解できるように学習する。|
|3rd practice test||100.0 points||Pass||September 06, 2020||Mock exam content After reviewing, retake the exam.|
|1st practice test||100.0 points||Pass||September 07, 2020||Implemented once a day as a target.|
|2nd practice test||100.0 points||Pass||September 08, 2020||Implemented once a day as a target.|
|3rd practice test||100.0 points||Pass||September 09, 2020||Implemented once a day as a target.|
|1st practice test||100.0 points||Pass||September 10, 2020||Implemented once a day as a target.|
|2nd practice test||100.0 points||Pass||September 11, 2020||Implemented once a day as a target.|
|3rd practice test||097.5 points||Pass||September 12, 2020||One question is wrong due to a general mistake.|
Terms of Service has the following description, so I cannot mention the contents. Also, regarding the email address, it is safer to use an address that does not cause any problems even if the email is delivered.
- The user shall not infringe or may infringe the copyright of our work, such as copying, public transmission, transfer, adaptation and translation.
- The user agrees that we will send you an email regarding recruitment. We will not disclose any standards for email delivery related to recruitment.
|Examination hall||「Oddessey ID & PasswordAnd don't forget the "designated ID".[Oddessey ID & Password]If you can't remember, the staff will allow you to bring in a sticky note. Other test takers besides me also received sticky notes.
I was allowed to go to the bathroom on the way during other tests, but this test was based on the premise that I would end when I went out, even if I was ill. Many times before the test started, I was prompted to go to the bathroom in advance, which put pressure on me. I was nervous when I took the exam if my stomach hurts on the way.
|Impressions||The content of the problem cannot be described by the rules. It took about 30 minutes to solve 40 questions. However, it was dangerous because I was worried about answering about 12 out of 40 questions. 28 questions because it passed with 700 points or more(28 x 25 points = 700 points)Was a last-minute line that needed to be answered correctly and could be mistaken. However, all I didn't understand was memorization problems, and I couldn't help taking time, so I gave up and ended the test.
It is a point assuming the same Python beginner as me.
There are two qualifications that you should obtain in advance. There are the following two. If you have reached the level where you can acquire these two, I think that you will understand the qualification study of data analysis quickly.
「Gtest合格者による受験に向けた心構えについて（2020年7月合格）I mentioned the article in ", but it is a qualification that can be obtained in a study period of about one month.
|Python 3 Engineer Certification Basic Exam||IfyouarenewtoPython,thisisdefinitelythewaytogo.IthinkthedataanalysistestisessentialbecauseitrequiresthatyouknowthebasicsofPython.
「HowtostudyPython3engineercertificationbasicexambyPythonbeginner(passedinAugust2020)I mentioned the article in ", but it is a qualification that can be obtained in a study period of about one month.
As mentioned earlier, "New data analysis textbook using Python (official) The official textbook is a difficult material for beginners to study first. Therefore, I think it is efficient to study in advance so that the contents of this teaching material will be in your mind.
Udemy "[Learn in one day] Python data analysis for engineers" that I used for learning was very effective, so please I recommend it. The video will teach you how to install Anaconda, how to use Spyder and Jupyter notobook, how to operate Python, and how to use libraries such as Numpy in an easy-to-understand manner. When I read the official textbook after learning this, I was able to understand it to some extent. I think it will lead to more efficient learning and shorter time. Regarding this pre-learning material, I think that the material that suits each individual is different, so if it does not suit you, please try to find the material that suits you.
① When purchasing teaching materials at "Udemy), please purchase them at the time sale held about once every three months. This teaching material usually costs 10,800 yen, but the price drops to around 1,400 yen only during the time sale period.
(2) Some videos are released as samples in the "Course content" part of the teaching materials. Please read this content and purchase it if you think it suits you. If it doesn't suit you, I think you should look for other teaching materials.
As a preparation method for the production test, I think it is effective to carry out until you thoroughly understand "■ PRIME STUDY practice test". I will. It is important to study until you understand not only the correct answer, but also why the other options are wrong.
In addition, as a supplement to the mock questions when studying, all three mock questions are classified as follows. For example: 1. Problems related to the role of data engineer are always 1 and 2. If you know this premise, you can compare the contents of each mock question and find out where in the official text you can find the description.
|Chapter number||title||The number of questions||Problem rate||Mock question question number|
|1||Role of data engineer||2||5.00%||1,2|
|2.1||Execution environment construction||1||2.50%||3|
|3.1||Basic knowledge for reading mathematical formulas||1||2.50%||8|
|3.4||Probability and statistics||2||5.00%||12,13|
|5||application:Data collection and processing||0||0.00%|
However, as stated in my test results, there is a possibility that you will pass the exam with just the right score. From my reflection, I think it's better to organize functions and methods into EXCEL or something and understand them to some extent. The more content you remember, the better your score will be and the more you will be able to pass.
I will describe what I was interested in while implementing the official text.
There is a procedure to install Python in "2.1 Execution environment construction". However, if you already have Anaonda installed, you do not need to install it. "In the Anaconda environment, pip is also available, so it is possible to use pip to add these packages. However, in rare cases, the environment built with the conda command may be corrupted." There is also a description that the Anaconda environment will be destroyed at worst, so let's stop. Even if you leave Anaconda as it is, there is no problem in the range up to Chapter 4 of the test range.
When I buy the kindle version, paste the syntax and try to run it in python, I immediately get the following message:
However, posting example sentences by hand takes an enormous amount of time. What should not be overlooked here is the sample data that describes the download site in the text. If you upload the folder "notobooks" downloaded here directly under the designated folder of "jupyter", you can open the page with the syntax of Chapter 4. Therefore, you can execute it immediately and check the operation.
You need to install a module called "pydotplus" at "● Decision Tree" in "4.4.2 Classification". There was no description of the installation procedure by Anaconda, so I will describe it. Before doing this, let's stop "Jupyter" once. When you start "Jupyter" after the installation environment, the "pydotplus" module is available. ① Start Wincows "Start" button> Anaconda3 (64-bit)> Anaconda Navigator (anaconda3). ② Select "Environments". ③ Select "All". ④ Select "Update index" to update the index.
An error will occur even if you write it according to the syntax with "scikit-learn". This is an error because some classes are obsolete. I was also in trouble, but there is a countermeasure on the page "I tried to deal with the scikit-learn warning I encountered during copying". ,It was helpful.
It was great to be able to study the specific behavior of libraries such as Numpy, Pandas, Matplotlib, and scikit-learn that use Python. Even if I studied the library name in the G test, I could not understand the specific part such as what kind of processing is actually done. That question was resolved through this qualification study.
I hope this article will serve as a reference for those who are taking the "Python3 Engineer Certification Data Analysis Exam".