In phdornever/Bjj: Analyse information data of jiu-jitsu trainners in China

Motivation

I am always a big fan of jiu-jitsu,a special martial art. Nowadays, jiu-jitsu is still at its beginning of spreading in China. Still very few people know it , let alone train it. Everything is like chaos, because jiu-jitsu just grows agressively, I have a strong feeling it will furish in China as well as Tae koon doo, another martial art travelled from Korea. Therefore, I think it is a good time to collect some data and record this time. I need to know these people's age, gender,etc. Information about what kind of people they are, how many times they train a week..... And I have a sense that in China, jiu-jitsu is kind of exclusive to certain groups of people, who have high salary, are young and not busy working, love martial arts, be willing to train for a long time. I hope I can prove my guess through my package and my data, though they are not complete yet.

what I want to learn

I want to learn the mean age of jiujitsu trainner in China, thus I use sample to estimate it. I also use the sample injure rate to represent population injure rate.I calculate mean spending of each place in my sample to learn how much people spend in jiujitsu in different places.

The data I collect

I collect the data of jiu-jitsu trainers. It includes their age(under 18,18 to 25 etc), gender, how long they've been training(1 year or 1.5 year), their money spent on jiu-jitsu(for example 8000 Chinese yuan per year), and their training frequency(for instance, 1 to 3 times a week or 3 to 5 times a week), competition plan(for example, 1 times a year).

Where I get the data?

I use a popular China websit to create a survey form and send it to as many jiu-jitsu groups as I can.

Survey link：https://www.wjx.cn/jq/16656266.aspx

Experimental Design: Nowadays, not many people in China know jiu-jitsu, most of our trainners are concentrated in well developed cities, like Shanghai, Chengdu, Beijing....So I think my data is representative because I collected this data from many big famous jiu-jitsu clubs in those cities. Besides, I also collect data from smaller cities such as Wenzhou....I only ask people simple questions so that they can answer easily with yes, or no or some certain number. Thus I think my measure is valid. OK I take my words back, this is my first time making survey and analyses data, I am not sure how to use my statistical method to my data..My data is not so randomized because I only send survey to certain clubs. And the potential confront would be that I did not send my survey to a better representative group, I only let my friends send it to others as possible as they can, I am not sure they did it or not.

What is this package for?

This packge is created for plotting information in the jiu-jitsu data, and doing some caculations with the data. It deals with a lot variables in the data collected by the survey.

By using this package, we can easily get data from googlesheets through a single function. We can also use simplified function building some dataframe and analyse special parts of data, rather than using repetitions or create functions temporarily.

We can also use one function to caculate the mean of money spend per year. Getting injure rate of this sport.

As time move on, I will keeping adding new abilities for this package.

First, we should ask googlesheets permission code.

library(BjjAnalyse)
authoriz()

Examples: Get collection of data from googlesheets, and put it in object dat.

library(knitr)
dat <- read_data()
kable(dat[1:10,1:10])

Example: plot a hist graph about age and trainning frequency by using age_freq().

library(BjjAnalyse)
age_freq(dat)

Example: People might want to know the injure rate of this kind of sport.We can use inj_rate()

inj_rate(dat)

Example: If I want to know the average amount money of each location,I can use loc_mean() function.

library(BjjAnalyse)
loc_mean(dat)[1:5]

Example: Get a pie graph of the percentage of each belt rank of self rank in all, in the example we can see for instance, people who is white belt is 69.51% of all.

library(BjjAnalyse)
self_rank(dat)

Results of my analysis: In China jiu-jitsu is a kind of expensive sports, people cost more than 8000 chinese Yuan per year even in poor province like anhui, hebei. Most of jiu-jitsu trainers in China are freshmen. Jiu-jitsu is dangerous, trainers have high risk to get injured. Jiu-jitsu is not a friendly sport towards old people. old people tend to train much less than youngers.

What I learned from this homework and my limitations: I learned how to use version control to develop my package in a long period. However, I am still not good at git, I need more practices. The limitation of my analysis is that the data is ralatively small and my data-collect-way is not very formal, both lead to a limited conclusion of my analysis, and it is also hard to draw more complex things out of my data. My analysis is very simple and doesn't contain many useful estimators like variance. I could perform a linear regression on my data, that might make my analysis more complete. However the reason I didn't do that is that I didn't think it will make sense to do a linear regression to my data.

What will I do differently next time. I will use more ways of statistical methods, such as t-test, linear regression, which is what I just learned this semester. About data collecting, I will try to collect data that can be expressed in numbers, like age, height, not some yes or no data.

phdornever/Bjj documentation built on Aug. 29, 2019, 12:40 p.m.