fix minor typos

alexd106 · Mar 26, 2024 · 396220c · 396220c
1 parent cc11b5e
commit 396220c
Show file tree

Hide file tree

Showing 18 changed files with 89 additions and 87 deletions.
diff --git a/exercise_solutions/graphical_data_exploration_exercise_solutions.R b/exercise_solutions/graphical_data_exploration_exercise_solutions.R
@@ -1,4 +1,4 @@
-## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## loyn <- read.table("./data/loyn.txt", header = TRUE,
 ##                    stringsAsFactors = TRUE)
 ## str(loyn)
@@ -13,14 +13,14 @@
 ## loyn$FGRAZE <- factor(loyn$GRAZE)
 
 
-## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## table(loyn$FGRAZE)
 ## 
 ## # or use xtabs function
 ## xtabs(~ FGRAZE, data = loyn)
 
 
-## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # mean abundance of birds for each level of FGRAZE
 ## tapply(loyn$ABUND, loyn$FGRAZE, mean, na.rm = TRUE)
 ## 
@@ -31,7 +31,7 @@
 ## tapply(loyn$ABUND, loyn$FGRAZE, summary, na.rm = TRUE)
 
 
-## ----Q7a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q7a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # first split the plotting device into 2 rows
 ## # and 3 columns
 ## par(mfrow = c(2,3))
@@ -45,7 +45,7 @@
 ## dotchart(loyn$GRAZE, main = "Grazing levels")
 
 
-## ----Q7b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q7b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # A fancier version of a dotplot - just for fun!
 ## Z <- cbind(loyn$ABUND, loyn$AREA, loyn$DIST,
 ##            loyn$LDIST,loyn$YR.ISOL,loyn$ALT,
@@ -68,7 +68,7 @@
 ##         ylab = "Order of the data from text file")
 
 
-## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # There appears to be two unusually large forest patches compared to the rest
 ## # Also one potentially large distance in DIST
 ## # One option would be to log10 transform AREA, DIST
@@ -81,7 +81,7 @@
 ## str(loyn)
 
 
-## ----Q9a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q9a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # Vanilla pairs plot
 ## 
 ## pairs(loyn[,c("LOGAREA", "LOGDIST", "LDIST",
@@ -95,7 +95,7 @@
 ## pairs(explan_vars)
 
 
-## ----Q9b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q9b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # And with correlations in the upper panel
 ## 
 ## # first need to define the panel.cor function
@@ -115,23 +115,23 @@
 ##       upper.panel = panel.cor)
 
 
-## ----Q10a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE---------------------------------------------------------------------
+## ----Q10a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
 ## pairs(loyn[,c("ABUND","LOGAREA","LOGDIST", "LDIST",
 ##     	"YR.ISOL","ALT","GRAZE")],
 ##       	lower.panel = panel.cor)
 
 
-## ----Q10b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE---------------------------------------------------------------------
+## ----Q10b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
 ## plot(loyn$LOGAREA, loyn$ABUND, xlab = "log area", ylab = "bird abundance")
 
 
-## ----Q11a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE---------------------------------------------------------------------
+## ----Q11a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
 ## # Interaction between LOGAREA and FGRAZE?
 ## # Do the slopes look similar or different?
 ## coplot(ABUND ~ LOGAREA | FGRAZE, data = loyn)
 
 
-## ----Q11b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE---------------------------------------------------------------------
+## ----Q11b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
 ## # Fancier version of the above plot
 ## # with a line of best fit included just for fun
 ## coplot(ABUND ~ LOGAREA | FGRAZE,

diff --git a/exercise_solutions/linear_model_1_exercise_solutions.R b/exercise_solutions/linear_model_1_exercise_solutions.R
@@ -1,12 +1,12 @@
-## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## loyn <- read.table("data/loyn.txt", header = TRUE,
 ##                    stringsAsFactors = TRUE)
 ## str(loyn)
 ## 
 ## # 67 observations and 8 variables (from str())
 
 
-## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # let's first log10 transform the AREA variable due to the couple
 ## # of unusually large forest area values (check you data exploration)
 ## # to remind yourself
@@ -23,7 +23,7 @@
 ## loyn.lm <- lm(ABUND ~ LOGAREA, data = loyn)
 
 
-## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # ANOVA table
 ## anova(loyn.lm)
 ## 
@@ -36,7 +36,7 @@
 ## # from 0)
 
 
-## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## summary(loyn.lm)
 ## 
 ## # the estimate of the intercept = 10.4
@@ -46,7 +46,7 @@
 ## # ABUND = 10.40 + 9.78 * LOGAREA
 
 
-## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## 
 ## # the null hypothesis for the intercept is that the intercept = 0
 ## # the p value is very small (certainly less than the (not so magic) 0.05)
@@ -60,14 +60,14 @@
 ## # LOGAREA and ABUND).
 
 
-## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## summary(loyn.lm)
 ## 
 ## # The multiple R-squared value is 0.588 and therefore 58.8% of
 ## # the variation in ABUND is explained by LOGAREA
 
 
-## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # first split the plotting device into 2 rows and 2 columns
 ## par(mfrow = c(2,2))
 ## 
@@ -102,7 +102,7 @@
 ## # log transformation things look ok.
 
 
-## ----Q9, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q9, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # to predict bird abundance if AREA == 100
 ## 
 ## # if you log base 10 transformed the AREA variable
@@ -112,7 +112,7 @@
 ## # not log10()
 
 
-## ----Q10, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q10, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # need to create a dataframe object with a column of LOGAREA values to predict from.
 ## # note you need to call the column the same as used in the model
 ## my.data <- data.frame(LOGAREA = seq(from = min(loyn$LOGAREA),
@@ -124,7 +124,7 @@
 ## pred.vals <- predict(loyn.lm, newdata = my.data)
 
 
-## ----Q11, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q11, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # plot the lines on the plot. The x values are the new LOGAREA values from the my.data
 ## # dataframe, the predicted values are from pred.vals
 ## plot(loyn$LOGAREA, loyn$ABUND, xlab = "Log10 Patch Area",
@@ -148,7 +148,7 @@
 ##     theme_classic()
 
 
-## ----Q12, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q12, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## pred.vals.se <- predict(loyn.lm, newdata = my.data, se.fit = TRUE) # note the use of the se.fit argument
 ## 
 ## # check out the structure of pred.vals.se
@@ -188,7 +188,7 @@
 ##     theme_classic()
 
 
-## ----Q13, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q13, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # back transformed LOGAREA and 95% confidence intervals
 ## 
 ## # re-plot but this time use the original untransformed AREA variable

diff --git a/exercise_solutions/linear_model_2_exercise_solutions.R b/exercise_solutions/linear_model_2_exercise_solutions.R
@@ -1,18 +1,18 @@
-## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## loyn <- read.table("data/loyn.txt", header = TRUE,
 ##                    stringsAsFactors = TRUE)
 ## str(loyn)
 
 
-## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # create factor GRAZE as it was originally coded as an integer
 ## loyn$FGRAZE <- factor(loyn$GRAZE)
 ## 
 ## # check this
 ## class(loyn$FGRAZE)
 
 
-## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## boxplot(ABUND ~ FGRAZE, xlab = "Grazing level", ylab = "Bird abundance", data = loyn)
 ## 
 ## # mean bird abundance for each level of FGRAZE
@@ -26,11 +26,11 @@
 ## # levels 2,3 and 4.
 
 
-## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## birds.lm <- lm(ABUND ~ FGRAZE, data = loyn)
 
 
-## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## anova(birds.lm)
 ## 
 ## # null hypothesis : There is no difference in the mean bird abundance between the
@@ -43,7 +43,7 @@
 ## # of grazing intensity (F_4,62 = 14.98, p < 0.0001)
 
 
-## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## summary(birds.lm)
 ## 
 ## # Here the intercept (baseline) is the mean abundance of birds for FGRAZE level 1.
@@ -73,7 +73,7 @@
 ## # significantly different from graze level 5.
 
 
-## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # Set FGRAZE level 2 to be the intercept
 ## 
 ## loyn$FGRAZE <- relevel(loyn$FGRAZE, ref = "2")
@@ -97,12 +97,12 @@
 ## # The intercept is now FGRAZE level 4, we can now compare between levels '4 and 5'
 
 
-## ----Q9, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
+## ----Q9, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE------------------------------------------------------------------------
 ## # The multiple R-squared value is 0.491 and therefore 49.1% of
 ## # the variation in ABUND is explained by FGRAZE
 
 
-## ----Q10, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE----------------------------------------------------------------------
+## ----Q10, eval=SOLUTIONS, echo=SOLUTIONS, collapse=TRUE-----------------------------------------------------------------------
 ## # first split the plotting device into 2 rows and 2 columns
 ## par(mfrow = c(2,2))
 ## 
@@ -140,7 +140,7 @@
 ## # generalised least squares. This is not something we will do on this course but will cover in a more advanced statistics course!
 
 
-## ----Q11a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE--------------------------------------------------------------------
+## ----Q11a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
 ## # Using the gplots package, you may need to install this package first
 ## # install.packages('gplots')
 ## 
@@ -150,7 +150,7 @@
 ##   ylab = "bird abundance", data = loyn, connect = FALSE)
 
 
-## ----Q11b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE--------------------------------------------------------------------
+## ----Q11b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
 ## # Using the effects package, you may need to install this package first
 ## # install.packages('effects')
 ## 
@@ -159,7 +159,7 @@
 ## plot(loyn.effects,"FGRAZE", lty = 0)
 
 
-## ----Q11c, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE--------------------------------------------------------------------
+## ----Q11c, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
 ## # and finally using old faithful the predict function and base R graphics
 ## # with the segments function
 ## 
@@ -174,7 +174,7 @@
 ## segments(1:5, pred.vals$fit, 1:5, pred.vals$fit + 1.96 * pred.vals$se.fit)
 
 
-## ----Q11d, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE--------------------------------------------------------------------
+## ----Q11d, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
 ## # using old faithful the predict function and base R graphics
 ## # with the arrows function
 ## 

diff --git a/exercise_solutions/linear_model_3_exercise_solutions.R b/exercise_solutions/linear_model_3_exercise_solutions.R
@@ -1,16 +1,16 @@
-## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q2, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## loyn <- read.table("data/loyn.txt", header = TRUE,
 ##                    stringsAsFactors = TRUE)
 ## str(loyn)
 
 
-## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q3, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## loyn$LOGAREA <- log10(loyn$AREA)
 ## # create factor GRAZE as it was originally coded as an integer
 ## loyn$FGRAZE <- factor(loyn$GRAZE)
 
 
-## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q4, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## coplot(ABUND ~ LOGAREA | FGRAZE, data = loyn)
 ## 
 ## # or
@@ -27,14 +27,14 @@
 ## #   need to test this with a model.
 
 
-## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q5, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## birds_inter1 <- lm(ABUND ~ FGRAZE + LOGAREA + FGRAZE:LOGAREA, data = loyn)
 ## 
-## # Or use the 'shortcut' - its is equivalent to the model above
+## # Or use the 'shortcut' - it's equivalent to the model above
 ## # birds.inter.1 <- lm(ABUND ~ FGRAZE * LOGAREA, data = loyn)
 
 
-## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q6, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## # first split the plotting device into 2 rows and 2 columns
 ## par(mfrow = c(2,2))
 ## 
@@ -58,7 +58,7 @@
 ## # Leverage plot (all < 0.5).
 
 
-## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q7, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## anova(birds_inter1)
 ## 
 ## # The null hypothesis is that there is no significant interaction between
@@ -76,7 +76,7 @@
 ## # dependent on the value of the other variable.
 
 
-## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE----------------------------------------------------
+## ----Q8, eval=SOLUTIONS, echo=SOLUTIONS, results=SOLUTIONS, collapse=TRUE-----------------------------------------------------
 ## summary(birds_inter1)
 ## 
 ## # (Intercept)
@@ -134,7 +134,7 @@
 ## # LOGAREA and FGRAZE as single explanatory variables.
 
 
-## ----Q9a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
+## ----Q9a, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE----------------------------------------------------------------------
 ## par(mfrow= c(1, 1))
 ## plot(ABUND ~ LOGAREA, data = loyn, col = GRAZE, pch = 16)
 ## # Note: # colour 1 means black in R
@@ -209,7 +209,7 @@
 ##  lwd = c(1, 1, 1))
 
 
-## ----Q9b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
+## ----Q9b, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE----------------------------------------------------------------------
 ## # Okay, that was a long-winded way of doing this.
 ## # If, like me, you prefer more compact code and less risks of errors,
 ## # you can use a loop, to save repeating the sequence 5 times:
@@ -231,7 +231,7 @@
 ##  lwd = c(1, 1, 1))
 
 
-## ----Q9c, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE---------------------------------------------------------------------
+## ----Q9c, eval=SOLUTIONS, echo=SOLUTIONS, collapse=FALSE----------------------------------------------------------------------
 ## # install.packages('ggplot2', dep = TRUE)
 ## library(ggplot2)
 ##