Statistics with R - Myanmar

Probability Distributions!
Learn Statistics with these Best Resources-> https://www.mltut.com/best-statistics-courses-on-coursera/

Image Credit-> DataInterview

Ballon plot by using gplots package

Categorical variables တွေကို Contingency table (နှစ်ဖက်မြင်ဇယား) တည်ဆောက်ပုံနှင့် တည်ဆောက်ထားတဲ့ နှစ်ဖက်မြင်ဇယား ကို graphical display နဲ့ ပုံဖေါ်လိုပါကသုံးလို့ရပါတယ်၊
factoextra package ကို library function သုံးပြီး R section ထဲကိုခေါ်ပါ။
ဒီ package မှာပါတဲ့ housetasks ဆိုတဲ့ data ကိုသုံးမှာမို့ပါ။ data ရဲ့သဘောတရားကိုသိအောင် View() နဲ့ကြည့်ပါမယ်။
data structure ကိုသိအောင် str() function သုံးပြီးကြည့်တာ observation ၁၃ ခု၊ variable ၄ခု ပါတဲ့ data.frame ဖြစ်တာကိုတွေ့ရပါတယ်။
row မှာ laundry, Main_meal, Dinner, Breakfast, etc., အစရှိတဲ့ observations 13 ခု နဲ့ column မှာ variables ၄ ခု ဖြစ်တဲ့ Wife (ဇနီးတာဝန်ယူ)၊ Alternating (တလှည့်စီတာဝန်ယူ)၊ Husband (အမျိုးသားတာဝန်ယူ)၊ Jointly (အတူတူတာဝန်ယူ) တို့ဖြစ်ပြီး တန်ဖိုးတွေက အိမ်အလုပ်တွေကို တာဝန်ယူမှုကြိမ်နှုန်းတွေဖြစ်ပါတယ်။

library (factoextra)
View (housetasks)
str (housetasks)


ပထမဆုံး data.frame ကို table အဖြစ် ပြောင်းပါတယ်

housetasks

Learn Statistics with these Online Resources-> https://www.mltut.com/best-statistics-courses-on-coursera/

ML Algorithms where Data Normalization is needed and not.
Learn Machine Learning with these Online Courses- https://www.mltut.com/best-advanced-machine-learning-courses/

Image Credit-> Avi Chawla

𝐔𝐒𝐄𝐅𝐔𝐋 𝐖𝐄𝐁𝐒𝐈𝐓𝐄𝐒 𝐅𝐎𝐑 𝐑𝐄𝐒𝐄𝐀𝐑𝐂𝐇

𝐆𝐨𝐨𝐠𝐥𝐞 𝐒𝐜𝐡𝐨𝐥𝐚𝐫
http://scholar.google.com/

It is one of the widely used academic resources for researchers and scholars, a friendly free academic search engine that indexes academic information from various online web resources. The Google Scholar lists information across an array of academic resources, mostly are peer-reviewed. It works in the same manner as Scirus.

𝐃𝐢𝐫𝐞𝐜𝐭𝐨𝐫𝐲 𝐨𝐟 𝐎𝐩𝐞𝐧 𝐀𝐜𝐜𝐞𝐬𝐬 𝐉𝐨𝐮𝐫𝐧𝐚𝐥𝐬 (𝐃𝐎𝐀𝐉)
http://www.doaj.org/

This offers a huge range of topics within scientific areas of study. It is among the richest sources of scholarly database with over 8,000 journals available on different topics. All the journals are thoroughly peer-reviewed.

𝐄𝐝𝐮𝐜𝐚𝐭𝐢𝐨𝐧 𝐑𝐞𝐬𝐨𝐮𝐫𝐜𝐞𝐬 𝐈𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐂𝐞𝐧𝐭𝐞𝐫 (𝐄𝐑𝐈𝐂)
https://eric.ed.gov/

It allows you to search by topics related to education. Links lead to other sites, where you may have to purchase information, but you can search for full-text articles only.

𝐒𝐩𝐫𝐢𝐧𝐠𝐞𝐫
https://www.springernature.com
https://link.springer.com

It helps researchers find relevant resources, providing links to full articles, where available, or links to where the book or article can be accessed for a fee.

𝐄𝐁𝐒𝐂𝐎
https://www.ebsco.com/open-access

It provides access to over 172,000 dissertations from 1902 to the present in a variety of subjects. Full texts are provided when available.

𝐂𝐎𝐑𝐄
https://core.ac.uk/

CORE aims to aggregate all open access research and facilitate unrestricted access to research. CORE harvests research papers from data providers from all over the world including institutional and subject repositories, open access and hybrid journal publishers. It contains 192,131,186 open access articles, from thousands and over tens of thousands of journals, collected from 10,079 data providers around the world.

𝐒𝐞𝐦𝐚𝐧𝐭𝐢𝐜 𝐒𝐜𝐡𝐨𝐥𝐚𝐫
https://semanticscholar.org

This easy-to-use resource uses artificial intelligence to efficiently sort through millions of science-related papers based on your search terms. According to the site, although some articles are behind paywalls, "the data [they] have for those articles is limited."

𝐉𝐒𝐓𝐎𝐑
https://www.jstor.org

Is a very well-known resource, offering access to journals, primary sources, and books in a variety of disciplines. You can register for a free account, but many of the articles are behind a paywall.

𝐏𝐇𝐋 𝐂𝐇𝐄𝐃 𝐂𝐨𝐧𝐧𝐞𝐜𝐭
https://phlconnect.ched.gov.ph/

It contains higher education materials in various formats that are useful for teaching, learning and research purposes.

𝐏𝐡𝐢𝐥𝐢𝐩𝐩𝐢𝐧𝐞 𝐄- 𝐉𝐨𝐮𝐫𝐧𝐚𝐥𝐬
https://ejournals.ph

It is a free online database that provide a huge of journals published in Philippines, covering the full range of academic disciplines.

𝐆𝐫𝐚𝐦𝐦𝐚𝐫𝐥𝐲
https://app.grammarly.com/

Is an app that automatically detects spelling, punctuation, grammar, word choice, and even your writing style mistakes. Just copy your text, paste on the Grammarly app and voila —time to check all the grammar mistakes you may have missed!

𝐏𝐮𝐫𝐝𝐮𝐞 𝐎𝐰𝐥
https://owl.purdue.edu/

It gives information about how to correctly cite references using various style guides such as APA, MLA, and the Chicago Manual of Style.

𝐆𝐨𝐨𝐠𝐥𝐞 𝐅𝐨𝐫𝐦
https://docs.google.com/forms/u/0/?tgif=d

Use Google Form to create create online forms and surveys with multiple question type.

Correlogram by using ggpairs function
R programming တွင် numerical variable ၄ခု (sepal length, sepal width, petal length, petal width) တို့၏ correlation ကို categorical variable ၁ခုဖြစ်တဲ့ species name အလိုက်ဆွဲထားခြင်းဖြစ်ပါတယ်၊
R မှာပါတဲ့ iris dataset ကိုအခြေခံထားပါတယ်။ command line တစ်ကြောင်းထဲနဲ့ density distributions of the variables, scatter plot of variables (eg. sepal length and sepal width, sepal length and petal width, etc.,), correlation coefficient and significant level တို့ကိုသရုပ်ဖေါ်ပေးနိုင်ပါတယ်။

# install.packages("GGally")
library(GGally)

ggpairs(iris,
columns = 1:4,
aes(color = Species,
alpha = 0.5))

# ggpairs(iris, က ကိုယ်သုံးမဲ့ Data frame ကိုရည်ညွှန်းပါတယ်
# columns =1:4, က iris data set ရဲ့ column 1 to 4 ကိုရည်ညွှန်းပါတယ်
# aes(color = Species, က cat. variable ဖြစ်တဲ့ species ကို color အလိုက် group ဖွဲ့ပီးဆွဲမယ်လို့ဆိုလိုတာပါ
# alpha = 0.5)) က color တွေကို transparency လုပ်မယ်လို့ရည်ညွှန်းပါတယ်။

The representation of the Data is Everything.

Simple, yet conveys a wealth of information.

Parametric test နဲ့ non-parametric test ဘယ်အချိန်မှာဘာသုံးမလဲ ဘာကွာသွားလဲဆိုတာ one sample test ကိုဥပမာပေးပြီးရှင်းထားပါတယ်၊
Parametric test ကို numerical dataများကို စစ်ဆေးရာတွင်သုံးသည်၊ Parametric test များမလုပ်မီ သက်ဆိုင်ရာ assumption များပြေလည်ရန်လိုသည်။ Non-parametric test ကို Categorical data များကို စစ်ဆေးရာတွင်သုံးသည်၊ numerical data များတွင်လည်း assumption များမပြေလည်ကလည်းသုံးကြသည်။

One-sample sign test
Non-parametric test အမျိုးအစားတစ်ခုဖြစ်သည်၊
sample data မှ median တန်ဖိုးကိုတွက်ယူပြီး သီအိုရီအရကြိုတင်သတ်မှတ်ထားသော median (theoretical median) တန်ဖိုးနှင့်နှိုင်းယှဉ်စစ်ဆေးခြင်းဖြစ်သည်။
Null hypothesis
H0: m=m0
Alternative hypothesis
two-tailed အတွက်ဆိုပါက H1: m≠m0 ဖြစ်ပြီး
one-tailed အတွက်ဆိုပါက H1: m>m0 (or) H1: m

Welcome to Statistics with R, Myanmar. On this page, we share knowledge about R programming and statistics. Additionally, we offer free webinars. Please follow and share this page's knowledge with your friends