Differences

This shows you the differences between two versions of the page.

--- c:ma:2018:schedule [2018/11/27 01:44] – hkimscil
+++ c:ma:2018:schedule [2018/12/17 10:01] (current) – [Week15 (Dec. 11, 14)] hkimscil
@@ Line 435: / Line 435: @@
 <WRAP half column>
 ===== Concepts and ideas =====
-[[:Regression]] [[:Multiple Regression]]
+Do the following
-[[:r:Linear Regression]] and [[:r:ANOVA]]
+<code>S1 <- c(89, 85, 85, 86, 88, 89, 86, 82, 96, 85, 93, 91,
-  - Introduction
+, 87, 94, 77, 87, 98, 85, 89, 95, 85, 93, 93,
-  - Performing Simple Linear Regression
+, 71, 97, 93, 75, 68, 98, 95, 79, 94, 98, 95)
-  - Performing Multiple Linear Regression
+S2 <- c(60, 98, 94, 95, 99, 97, 100, 73, 93, 91, 98,
-  - Getting Regression Statistics
+, 66, 83, 77, 97, 91, 93, 71, 91, 95, 100,
-  - Understanding the Regression Summary
+, 96, 91, 76, 100, 97, 99, 95, 97, 77, 94,
-  - Performing Linear Regression Without an Intercept
+, 88, 100, 94, 93, 86)
-  - Performing Linear Regression with Interaction Terms
+S3 <- c(95, 86, 90, 90, 75, 83, 96, 85, 83, 84, 81, 98,
-  - Selecting the Best Regression Variables
+, 94, 84, 89, 93, 99, 91, 77, 95, 90, 91, 87,
-  - Regressing on a Subset of Your Data
+, 76, 99, 99, 97, 97, 97, 77, 93, 96, 90, 87,
-  - Using an Expression Inside a Regression Formula
+, 88)
-  - Regressing on a Polynomial
+S4 <- c(67, 93, 63, 83, 87, 97, 96, 92, 93, 96, 87, 90,
-  - Regressing on Transformed Data
+, 90, 82, 91, 85, 93, 83, 90, 87, 99, 94, 88,
-  - Finding the Best Power Transformation (Box?Cox Procedure)
+, 72, 81, 93, 93, 94, 97, 89, 96, 95, 82, 97)
-  - Forming Confidence Intervals for Regression Coefficients
-  - Plotting Regression Residuals
+scores <- list(S1=S1,S2=S2,S3=S3,S4=S4)</code>
-  - Diagnosing a Linear Regression
-  - Identifying Influential Observations
+  * find means for each element in "scores" in a list format
-  - Testing Residuals for Autocorrelation (Durbin?Watson Test)
+  * find standard deviation for each element in "scores" in a data frame format
-  - Predicting New Values
+  * find variance  for each element in "scores" in a data frame format without using "var" function
-  - Forming Prediction Intervals
-  - Performing One-Way ANOVA
+<code>longdata<- c(-1.850152, -1.406571, -1.0104817, -3.7170704,
-  - Creating an Interaction Plot
+           -0.2804896, 0.9496313, 1.346517, -0.1580926, 1.6272786,
-  - Finding Differences Between Means of Groups
+           -2.4483321, -0.5407272, -1.708678, -0.3480616, -0.2757667,
-  - Performing Robust ANOVA (Kruskal?Wallis Test)
+           -1.2177024)</code>
-  - Comparing Models by Using ANOVA
+  * make "longdata" to a matrix whose size is  3 by 5
+  * name columns "trial1, trial2, . . . . trial5"
+  * name rows "subject1, subject2, subject3"
+  * get means for each subject
+  * attach the above data to the matrix data and name it "longtemp."
+  * get standard deviation for each trial
+  * attach the above data to the matrix data, "longtemp."
+<code>suburbs <- read.csv("http://commres.net/wiki/_export/code/r/data_transformations?codeblock=15", head=T, sep="	")</code>
+  * get subrubs data  as the above
+  * get population means by each state (listed in the data, suburbs)
+    * use aggregate and refer to the below e.g.
+<code>attach(Cars93)
+aggregate(MPG.city ~ Origin, Cars93, mean)</code>
+  * get population sum by each county with  tapply function.
+  * tapply(number, byfactor, function)
+  * how many counties are there?
+  * Use Cars93 data, get MPG.city mean by Origin.
+__Using pnorm, qnorm__
+pnorm : get proportion out of normal distribution whose characteristics are mean and sd
+<code>pnorm(84, mean=72, sd=15.2, lower.tail=FALSE)</code>
+  * What is the value of  the below?
+<code>pnorm(1)</code>
+  * How would you get 68, 95, 99% from pnorm
+      * use ?pnorm and see the default option
+  * generate 10 random numbers with runif function
+<code>year <- c(1900:2016)     # years in vector year
+world.series <- data.frame(year)</code>
+  * get 10 year samples out of world.series data with "sample" command
+  * how would you get the sample sample again latter?
+<code>pnorm(110, mean=100, sd=10)</code>
+  * What would be the result from the above?
+<code>library(MASS)       # load the MASS package
+tbl = table(survey$Smoke, survey$Exer)
+tbl                 # the contingency table</code>
+<code>summary(tbl)
+</code>
+  * read the above output and interpret
+  * what about the below one?
+<code>chisq.test(tbl)
+</code>
+see first [[:chi-square test]]
+see [[:r:chi-square test]] in r document space for more
+<code> library(MASS)
+ cardata <- data.frame(Cars93$Origin, Cars93$Type)
+ cardata
+</code>
+  * Can you say the types of cars are different by the Origins?
+<code>dur <- faithful$eruptions
+dur</code>
+  * make the above data into z-score (zdur).
+  * get mean of the zdur
+  * get sd of the zdur
+<code>
+set.seed(1123)
+x <- rnorm(50, mean=100, sd=15)
+</code>
+  * test x against population  mean 95.
+  * test x against population  mean 99.
+  * are they different from each other?
+  * what would you do if you want to see the different result from the  second one?
+<code>a = c(65, 78, 88, 55, 48, 95, 66, 57, 79, 81)
+> t.test(a, mu=60)
+	One Sample t-test
+data:  a
+t = 2.3079, df = 9, p-value = 0.0464
+alternative hypothesis: true mean is not equal to 60
+percent confidence interval:
+.22187 82.17813
+sample estimates:
+mean of x
+.2
+</code>
+  * find the t critical value with function qt.
+  * explain what happens in the next code
+  * read (or remind) what pnorm and qnorm do.
+<code>> s <- sd(x)
+> m <- mean(x)
+> n <- length(x)
+> n
+[1] 50
+> m
+[1] 96.00386
+> s
+[1] 17.38321
+> SE <- s / sqrt(n)
+> SE
+[1] 2.458358
+> E <- qt(.975, df=n-1)*SE
+> E
+[1] 4.940254
+> m + c(-E, E)
+[1]  91.0636 100.9441
+> </code>
+  * what's wrong with the below?
+<code>t.test(x)</code>
+<code>> mtcars</code>
+  * using aggregate, get mean for each trnas. type.
+  * compare the difference of mileage between auto and manual cars.
+    * use t.test (two sample)
+    * "use var.equal=T" option
+<code>a = c(175, 168, 168, 190, 156, 181, 182, 175, 174, 179)
+b = c(185, 169, 173, 173, 188, 186, 175, 174, 179, 180)
+</code>
+  * stack them into data c
+  * convert colnames into score and trans
+  * t.test score by trans with var.equal option true.
+  * aov test
+  * see  t.test t value, t = -0.9474 and F value,  F = ?
 </WRAP>
 <WRAP half column>
@@ Line 472: / Line 600: @@
 <WRAP half column>
 ===== Concepts and ideas =====
+ANOVA
+[[:r:oneway anova]]
+[[:r:twoway anova]]
+[[:r:linear regression]]
+[[:r:multiple regression]]
+[[:partial and semipartial correlation]]
+[[:statistical regression methods]]
+[[:sequential_regression]]
+[[:factor analysis]]
 Linear Regression and ANOVA
 http://commres.net/wiki/text_mining_example_with_korean_songs
-[[:temp|quiz 3 answer]]
 </WRAP>
 <WRAP half column>
 ===== Assignment =====
-  - 자신의 전공과 관심사에 맞는 아래의 테스트를 수행하기 위한 가설을 작성하시오.
-    - T-test
-    - F-test
-    - factorial f-test
-    - Simple regression
-    - Multiple regression
-  - 각 가설의 독립변인과 종속변인을 밝히고 이를 측정하는 방법에 대해서 논하시오.
-  - 가설과 관련이 있는 논문을 찾아서 (적어도 하나 이상씩) 관련 논문이 밝힌 것을 설명하고 자신의 가설과의 연관성을 논하시오.
-  - 각 가설에 필요한 데이터를 구한 후, 적절한 테스를 하시오 (r의 인풋과 아웃풋 필요).
-  - 테스트 결과를 논하시오.
 </WRAP>
 ====== Week15 (Dec. 11, 14) ======
 <WRAP half column>
-Group Presentation
+Final quiz
+Part I  (필기시험): NO open book.
+  * [[:correlation]]
+  * [[:regression]]
+  * [[:multiple regression]]
+  * [[:chi-square test]]
+  * [[:factor analysis]] - 이론적인 이해와 관련된 부분
+  * r 과 관련된 내용 중 통계에 대한 이해와 관련된 부분, 예를 들면
+    * t-test, ANOVA, Factorial  ANOVA output에 대한 이해
+    * regression, multiple regression output에 대한 이해 등
+Part II (r 실기시험): 교재와  R help만 허용
+  * [[:r:getting started]]
+  * [[:r:basics]]
+  * [[:r:navigating]]
+  * [[:r:input output]]
+  * [[:r:data structures]]
+  * [[:r:data transformations]]
+  * [[:r:probability]]
+  * [[:r:general statistics]]
+  * [[:r:t-test]]
+  * [[:r:anova]]
+  * [[:r:linear regression]]
+  * [[:r:multiple regression]]
+    * [[:partial and semipartial correlation]]
+    * [[:statistical regression methods]]
 </WRAP>
 <WRAP half column>
 </WRAP>
-<WRAP half column>
 ====== Week16 (Dec. 18, 21) ======
-Group Presentation
+<WRAP half column>
 __**Final-term**__
 </WRAP>