The beauty contest in 2021 has just ended. In the past few days, the number of visits to my blog has soared. It seems that there are many comrades participating in the competition.
Before the beauty contest, I made some study notes on the use of digital and analog algorithms and latex. The end of the game also means that the first phase of the winter vacation is over. It happens to be a summary here and also express some of my unique opinions.
The composition of our team: When we
first learned about mathematical modeling, others usually said to find teammates, one modeler, one programmer, and one essayist in a team. The division of labor is clear and reasonable. However, our team is a bit special. In the first three teams, no one has a clear expertise.
During the game, our team is arranged in this way, a tough player, responsible for continuously advancing the progress, I am in this role. The D question we chose has a total of 7 questions, and each question has many small questions (it should be the most questionable in the past years). I will draft my understanding of the question and my first opinion and make a draft. Some programming calculations; another modeling hand, with the strongest mathematics ability, he is responsible for the core topics, combined with my drafts to construct the core model; there is also a generalist who can play around, help with both sides, and do some basic programming calculations. There are pictures.
How do we choose the topic:
This time the topic is basically determined by me. I will publish the topic at 6 o'clock in the morning on the 2.5th. After watching for an hour, I will basically set the D question, although we are also familiar with network science and operations research. Nothing to understand.
Take a look at question
A this time : Fungi. Biological issues, and data is difficult to obtain, decisively give up.
Question B: Extinguish wildfires. At first glance, it looks like a cellular automata problem, but later it is related to drones and communications. It is somewhat complicated.
Question C: Rumors about wasps. This year's C question is very special. I gave the data set and more than 600 M files more than ten days ago. Before the competition, many people suspected the need to refine alchemy. The file has been encrypted, and a password is required to open it. I guessed someone would crack the code before, maybe I could solve the problem in advance. But the final password is: Af6SP7rdm33PxPJmDb4wZq7cw, such a long string is estimated that no one can brute force it through 10 days. After unpacking the file, I found that it was a lot of pictures of wasps. For me, it was a bit nauseating. I won’t put pictures here to pollute the eyes.
Question D: The influence of music. There is nothing special about the meaning of the question. The most important thing is that it has a data set. Although it is a bit large, except for the C and D questions, you need to collect data by yourself. This is why I choose it.
Question E: Re-optimize the food system. Question F: Check the pulse and temperature of higher education. Both questions feel a little abstract, especially question F, where quantitative data are somewhat difficult to obtain.
So in the end, we chose question D.
The arrangement of the question time:
a total of four days, we arranged like this:
Day 1: Choose the question, and complete the first question, the second question, in one day we will achieve half of the third question, the progress should be counted It’s
coming. The second day: After discussion, it was found that the first question has a radiation effect on the following questions. Therefore, we modified the model of the first question, and the progress basically advanced to the fifth question.
Day 3: The next two questions have a bit of language. The nature of the modeling questions is relatively fast. On the third day, I basically finished all the drafts. The typesetting was handed over to the modeling hand to sort out. He sorted out four questions in
one day. The fourth day (the last day): build The modeler read other papers and got new inspiration for the model of the first question, so I took time to modify it. I basically replaced him, sorted out the remaining questions, and drew some beautiful pictures with the generalist. .
A few days ago, we basically didn’t stay up late, but on the last day, the three of us found that the time was still a bit tight, so we didn’t submit it until 7 o’clock in the morning. In particular, this year’s new regulations for the Maison, the essay shall not exceed 25 pages. After finishing the layout, we found that there were 30 pages. We had to turn a picture of a picture into two side by side. The pictures were also shrunk and then reluctantly abandoned some language. Finally, 25 pages were filled.
Understanding of programming:
I see that many people looking for teammates have put forward high requirements on programming teammates. They must be proficient in C++, matlab, python... In fact, I think the most important thing is actually the processing of files! For some algorithms, it is not actually necessary to understand mathematical formulas, but only need to understand the application scenarios and input and output forms of different algorithms, the results, and some source code can be found. In this part, I also specially organized the mathematical modeling column. If you are interested, you can click on the portal https://blog.csdn.net/qq1198768105/category_10747260.html
Why it is said that the most important thing in programming is to process files. For example, in question D this time, an important piece of information (artist's genre information) in one table is in another table, and they need to be matched one-to-one, otherwise it is difficult to deal with. More than 4w data, manual completion is very unrealistic. And the three of us don't know how to process files. Later, I found out that a great god on Zhihu solved this problem. It's a life-saving medicine! Therefore, for programming, it is best to be able to master the pandas library and numpy library in python, and be able to batch process excel sheets.
Gains and software sharing: The
main gain of participating in the beauty contest is to discover so many weird scientific research software. During the preparation for the contest, I downloaded nearly 20 kinds of software. Here is also a summary to share.
Software for reading English papers:
Zhiyun document translation, you can explain it in one slide, and whoever uses it will know
Translation software:
DeepL is known as the most accurate translation software. If the resource download is slow, you can also go to my resources to download the installation package
Mathematical formula:
mathtype, not only can be exported graphically, but also latex language
mathpix: According to the image, directly OCR into a formula, especially easy to use when quoting other people's formulas
Paper editing software:
Checker, although I finally didn’t have time to use it this time
Draw a mind map:
MindMaster
Mindx
Draw a flowchart:
Visio
Data analysis software:
Origin, not only can analyze data, but also generate beautiful images
SPSS
Lingo, specializing in solving planning problems
matlab&pycharm, it is best to master these two main productivity (referring to understand basic grammar, will change other people's programs)
Drawing software:
this is the key, good pictures directly determine the person's first impression
Tableau: a relatively small minority but very useful software, the most important of which can be embedded foreign map, but I did not how used
Edraw map Note: Many pictures in this competition are drawn with it. It is a great software. There are many legends and templates in it.
WPS: Buy a one-day rice husk member. There are many beautiful pictures. It really costs money to make people stronger.
Excel : Excel actually hides a lot of potential. The latest version can also make 3D maps. Surprising
PPT: Many masters use this to draw
pictures. The drawing inside is very powerful. Unfortunately, I don’t know how to PS: With this basic, I can draw any map, but unfortunately I don't know much
Website for drawing special pictures:
Nightingale Rose Picture
Lessons learned :
1. Pay attention to the preservation of original data files and image files.
This time I got the data and started processing directly. I habitually created a sheet2 in the excel table. In the end, not only the original data was disrupted, but the image was also stored in a mess. Modifying an image requires repeated searches. .
2. Do it when you think of it, don't delay it to the end.
I thought that I needed to add a citation mark somewhere in the paper, so I wrote a note. I wanted to wait until the end of the reference to be marked again, but I almost forgot it in the end. .
3. The version control of the paper.
There are two people in our team who can write the paper. As a result, when the last revision is made, both of them modify their own version. The integration is more troublesome. In this case, it is best for three people to use a shared document to modify the content of the article. It is better to give the paper to one person.