diff --git a/assets/pptx/First-steps-with-R_day2_afternoon.pptx b/assets/pptx/First-steps-with-R_day2_afternoon.pptx
index 5c260af..92a5267 100644
Binary files a/assets/pptx/First-steps-with-R_day2_afternoon.pptx and b/assets/pptx/First-steps-with-R_day2_afternoon.pptx differ
diff --git a/assets/pptx/First-steps-with-R_day2_morning.pptx b/assets/pptx/First-steps-with-R_day2_morning.pptx
index e5ad684..d1f3141 100644
Binary files a/assets/pptx/First-steps-with-R_day2_morning.pptx and b/assets/pptx/First-steps-with-R_day2_morning.pptx differ
diff --git a/assets/pptx/First-steps-with-R_practicals_day1.pptx b/assets/pptx/First-steps-with-R_practicals_day1.pptx
index e428845..8f2a471 100644
Binary files a/assets/pptx/First-steps-with-R_practicals_day1.pptx and b/assets/pptx/First-steps-with-R_practicals_day1.pptx differ
diff --git a/assets/pptx/First-steps-with-R_practicals_day2.pptx b/assets/pptx/First-steps-with-R_practicals_day2.pptx
index 0c9c46a..8cf1c3c 100644
Binary files a/assets/pptx/First-steps-with-R_practicals_day2.pptx and b/assets/pptx/First-steps-with-R_practicals_day2.pptx differ
diff --git a/assets/pptx/exam_R.pptx b/assets/pptx/exam_R.pptx
index a0ace4c..18c8f23 100644
Binary files a/assets/pptx/exam_R.pptx and b/assets/pptx/exam_R.pptx differ
diff --git a/code/day1_code.R b/code/day1_code.R
index f8a5022..745b1ba 100644
--- a/code/day1_code.R
+++ b/code/day1_code.R
@@ -230,15 +230,19 @@ data(sleep)
 head(sleep)
 tapply(X=sleep$extra, INDEX=sleep$group, FUN=mean)
 
+## table
+table(snps$chr)
+table(snps$minor, snps$major)
+
 ##  adding rows
 snps_updated <- rbind(snps,
                       data.frame(chr=22, pos=1723369, minor="A", major="T"))
 snps_updated
 
 ## adding columns
-tested <- rep( c("yes",'no') ,  nrow(snps)/2)
-tested
-snps_mod <- cbind(snps, tested)
+majorGC <- snps$major %in% c("G","C")
+majorGC
+snps_mod <- cbind(snps, majorGC)
 snps_mod
 
 ## reshaping dataframe
diff --git a/code/day2_code.R b/code/day2_code.R
index 45479a4..46d643b 100644
--- a/code/day2_code.R
+++ b/code/day2_code.R
@@ -38,15 +38,41 @@ legend(x="bottomright",
        bg="gray90")
 
 
-## aparte - get data for practice
 
-## iris dataset
+## coloring 
 data(iris)
-head(iris)
 
-## generating random numbers
-rnorm(10)
-rnorm(10 , mean=10 , sd=2)
+plot(iris$Sepal.Length, 
+     iris$Sepal.Width , 
+     col = c('red','green','blue')[ iris$Species ] , 
+     pch=19)
+
+
+## coloring trick 
+iris$Species
+as.numeric(iris$Species)
+c('red','green','blue')[ c(1,1,1,2,2,2,3,3,3) ]
+
+c('red','green','blue')[ iris$Species ]
+
+## lines
+data(airquality)
+plot(airquality$Wind, airquality$Ozone, pch=20,
+     xlab="Wind (mph)",ylab="Ozone (ppb)")
+
+## horizontal and vertical lines
+abline(h=60, col="red", lty="dashed")
+abline(v=seq(3,21,3), col="grey", lty="dotdash")
+
+## tendency line
+abline(lm(airquality$Ozone ~ airquality$Wind),
+       col=2, lwd=2)
+
+legend("topright", legend= c("measures","fitted line"),
+       pch= c(20, NA), lty = c(0, 1), lwd=c(NA, 2),
+       col = c(1, 2), bg = "gray90")
+
+
 
 ## histogram
 x <- rnorm(10000)
@@ -92,28 +118,11 @@ points(thickness ~ status, data=Melanoma,
         col="blue", pch=19) #adds the actual data points to the plot
 
 
-## lines
-data(airquality)
-plot(airquality$Wind, airquality$Ozone, pch=20,
-     xlab="Wind (mph)",ylab="Ozone (ppb)")
-
-## horizontal and vertical lines
-abline(h=60, col="red", lty="dashed")
-abline(v=seq(3,21,3), col="grey", lty="dotdash")
-
-## tendency line
-abline(lm(airquality$Ozone ~ airquality$Wind),
-       col=2, lwd=2)
-
-legend("topright", legend= c("measures","fitted line"),
-       pch= c(20, NA), lty = c(0, 1), lwd=c(NA, 2),
-       col = c(1, 2), bg = "gray90")
-
 
 ## pairplot
 data(iris) #contains 4 measurements for 150 flowers from 3   species of iris (Iris setosa, versicolor and virginica)​
 pairs(iris[,1:4], main="Edgar Anderson's Iris Data", 
-        pch=21, bg=c("red", "green3", "blue")[iris$Species]) 
+        pch=19, col=c("red", "green3", "blue")[iris$Species]) 
 
 iris$Species
 as.numeric(iris$Species)
@@ -122,57 +131,19 @@ c('red','green','blue')[ c(1,1,1,2,2,2,3,3,3) ]
 
 c('red','green','blue')[iris$Species]
 
-## barplot
-
-course_data <- c(25, 35, 50, 100)
-barplot(course_data, main="Number of requests for R courses", 
-        names.arg=c("2011", "2012","2013", "2014"), 
-        col=c("yellow", "orange","red", "blue"))
-
-
-## more complex barplot
-df <- data.frame(year = c("2011","2012","2013","2014", 
-                          "2011","2012","2013","2014", 
-                          "2011","2012","2013","2014", 
-                          "2011","2012","2013","2014"),
-                 city =  c("A","A","A","A", "B","B","B","B", 
-                           "C","C","C","C", "D","D","D","D"),
-                 nb_requests_courses = c(30,36,50,98, 26,35,54,101,
-                                         28,38,51,105, 29,40,55,125))
-
-# Check what is inside
-df
-
-mean_nb  <- tapply(df$nb_requests_courses, df$year, mean)
-sd_nb    <- tapply(df$nb_requests_courses, df$year, sd)
-n_values <- tapply(df$nb_requests_courses, df$year, length)
-mean_nb
-sd_nb
-n_values
-
-
-mids <- barplot(mean_nb, 
-                xlab="year",ylab="Number of requests for courses", 
-                ylim=c(0,120),
-                col=c("yellow", "orange","red", "blue"),
-                cex=1.5, cex.axis=1.5, cex.main=1.5, cex.names=1.5,
-                main= "Number of requests for R")
-mids
-
-arrows(mids, mean_nb-sd_nb, # coordinates of lower point
-       mids, mean_nb+sd_nb, # coordinates of upper point
-       code=3,      # type of arrow: "head at both ends"
-       angle=90,    # angle between shaft and head of arrow
-       length=0.1,  # length of edges of arrow head
-       lwd=2)       # line width
-
-
-# Add text at the midpoints and at height 5 on the y-axis: number of observations​
-text(x=mids, y=5, paste("n =",n_values), cex=2)
 
 
+## par
 
+par(mfrow=c(1,2))
+plot(-90:90 , (-90:90)**2 )
+plot(-90:90 , (-90:90)**3 )
 
+par(mfrow=c(2,2))
+plot(-90:90 , (-90:90)**1 )
+plot(-90:90 , (-90:90)**2 )
+plot(-90:90 , (-90:90)**3 )
+plot(-90:90 , (-90:90)**4 )
 
 plot( rnorm(10) , rnorm(10) )
 par(col="red", pch=15)
@@ -205,40 +176,57 @@ dev.off()
 
 ### stats ###
 
-data(sleep)
-head(sleep)
+data(iris)
+# we limit the data to 2 species
+iris_f = iris[ iris$Species %in% c('versicolor','virginica') , ]
+iris_f$Species = factor(iris_f$Species)
+
+tapply( iris_f$Petal.Length ,iris_f$Species , mean )
+
+plen_versicolor = iris_f$Petal.Length[iris_f$Species=='versicolor']
+plen_virginica  = iris_f$Petal.Length[iris_f$Species=='virginica']
 
-summary(sleep)
 
 par(mfrow=c(2,2))
-hist(sleep$extra[sleep$group==1],
-      freq=FALSE, xlab="Drug 1",
-      main=" Extra sleep on drug 1")
-qqnorm(sleep$extra[sleep$group==1])
-qqline(sleep$extra[sleep$group==1])
-
-hist(sleep$extra[sleep$group==2],
-      freq=FALSE, xlab="Drug 2",
-      main=" Extra sleep on drug 2")
-qqnorm(sleep$extra[sleep$group==2])
-qqline(sleep$extra[sleep$group==2])
+hist(plen_versicolor,
+     xlab="versicolor",
+     main="petal length - versicolor")
+qqnorm(plen_versicolor)
+qqline(plen_versicolor)
+
+hist(plen_virginica,
+     xlab="virginica",
+     main="petal length - virginica")
+qqnorm(plen_virginica)
+qqline(plen_virginica)
 
 dev.off()
 
-boxplot(extra ~ group, data=sleep, col=c("orange", "pink"), ylab="Extra sleep", xlab="Drug received") 
-points(extra ~ group, data = sleep, col="black",pch = 19)
+shapiro.test( plen_versicolor )
+shapiro.test( plen_virginica )
+
+
+boxplot(Petal.Length ~ Species, data=iris_f, col=c("cornflowerblue", "pink"), 
+        ylab="petal length", xlab="species") 
+points( Petal.Length ~ Species, data=iris_f, col="black",pch = 19)
 
-t.test(sleep$extra[sleep$group==1],
-       sleep$extra[sleep$group==2])
+t.test(plen_versicolor,
+       plen_virginica)
 
-t.test(extra ~ group, data=sleep) #equivalent to the above 
+t.test(Petal.Length ~ Species, data=iris_f) #equivalent to the above 
 
 
-test_res = t.test(sleep$extra[sleep$group==1],
-                  sleep$extra[sleep$group==2])
+test_res = t.test(Petal.Length ~ Species, 
+                  data=iris_f)
 names(test_res)
 test_res[['p.value']]
 
+
+## paired t-test
+
+data(sleep)
+head(sleep)
+
 interaction.plot(response=sleep$extra, x.factor=sleep$group, 
                  trace.factor=sleep$ID, legend=FALSE, type="b", lty=1, pch=16, 
                  xlab="Drug received", ylab="Extra sleep")
@@ -297,15 +285,20 @@ class_data <- read.csv("course_datasets/class.csv")
 class_data$Gender = as.factor(class_data$Gender)
 #dataset* of 19 students' measurements
 summary(class_data)
-pairs(class_data)
+pairs(class_data[2:4] , col = class_data$Gender, pch=19 )
 
+# create model
 model_height_age <- lm(Height~Age, data=class_data)
 model_height_age
 
+plot(Height~Age, data=class_data)
+abline(model_height_age , col='red')
+
+residuals(model_height_age)
+
 ## check model assumptions
-par(mfrow=c(1,2))
-plot(model_height_age, which=1)
-plot(model_height_age, which=2)
+par(mfrow=c(2,2))
+plot(model_height_age)
 
 dev.off()
 ## plot model
@@ -322,3 +315,20 @@ abline(a=0,b=1,lty=2)
 
 summary( model_height_age )
 
+
+## model with 2 parameters
+
+
+# create model
+model2 <- lm(Height~Age+Gender, data=class_data)
+
+par(mfrow=c(2,2))
+plot(model2)
+dev.off()
+summary(model2)
+
+plot( Height~Age, col=c('turquoise','purple')[Gender], pch=19, data=class_data  )
+C = coef( model2 )
+abline(a=C[1],b=C[2] , col='turquoise',lwd=2)
+abline(a=C[1]+C[3],b=C[2] , col='purple',lwd=2)
+legend('bottomright',c('female','male'),pch=19,lwd=2,col=c('turquoise','purple'))
diff --git a/slides/First-steps-with-R_day2_afternoon.pdf b/slides/First-steps-with-R_day2_afternoon.pdf
index fb62831..3812d17 100644
Binary files a/slides/First-steps-with-R_day2_afternoon.pdf and b/slides/First-steps-with-R_day2_afternoon.pdf differ
diff --git a/slides/First-steps-with-R_day2_morning.pdf b/slides/First-steps-with-R_day2_morning.pdf
index d002696..95a87ff 100644
Binary files a/slides/First-steps-with-R_day2_morning.pdf and b/slides/First-steps-with-R_day2_morning.pdf differ
diff --git a/slides/First-steps-with-R_practicals_day1.pdf b/slides/First-steps-with-R_practicals_day1.pdf
index cda1e28..e6917c6 100644
Binary files a/slides/First-steps-with-R_practicals_day1.pdf and b/slides/First-steps-with-R_practicals_day1.pdf differ
diff --git a/slides/First-steps-with-R_practicals_day2.pdf b/slides/First-steps-with-R_practicals_day2.pdf
index 0c6a1a3..b4f1079 100644
Binary files a/slides/First-steps-with-R_practicals_day2.pdf and b/slides/First-steps-with-R_practicals_day2.pdf differ
diff --git a/slides/exam_R.pdf b/slides/exam_R.pdf
index 1286640..ed606c7 100644
Binary files a/slides/exam_R.pdf and b/slides/exam_R.pdf differ
diff --git a/solutions/R_practice10_solution.R b/solutions/R_practice10_solution.R
index c9bc42d..9dc898f 100644
--- a/solutions/R_practice10_solution.R
+++ b/solutions/R_practice10_solution.R
@@ -1,31 +1,95 @@
-
-# First Steps in R, Practice 10
-
-# 1) Load the package MASS using library(). (You may need to install it first).
-# Load the dataset Pima.tr using data(). Use ? to get an idea which variables it contains.
-
-library(MASS)
-data(Pima.tr)
-?Pima.tr
-
-
-#2) Hypothesis: Blood glucose level (glu) is associated with diastolic blood pressure (bp).
-#   Run a linear model to test the hypothesis.
-
-model_fit <- lm(bp~glu, data=Pima.tr)
-summary(model_fit)
-
-
-# the linear association between glucose and blood pressure is significant at the 0.05 alpha level
-# (p = 0.000115)
-
-# 3) Visualize the fit with a scatter plot and a trend line.
-par(mfrow=c(1,1))
-plot(Pima.tr$glu, Pima.tr$bp,  pch=20, xlab="blood glucose", ylab="blood pressure")
-abline(model_fit, col='red' , lwd=3)
-
-# 4) Check the assumption of normality of the residuals graphically.
-par(mfrow=c(2,1))
-plot(model_fit, which=1)
-plot(model_fit, which=2)
-# Looks quite ok
+# First Steps in R, Practice 10
+
+# Come back to the mice data-set stored in the "mice_data" data frame.
+
+mice_data <- read.table( "course_datasets/mice_data.csv" , header=T , sep=",",
+                         colClasses=c("factor", "factor", "numeric")) # define classes for columns  )
+
+mice_data$genotype <- factor(mice_data$genotype, levels=c("WT", "KO"))
+
+
+# 1)  Considering WT mice weight and KO mice weight separately, check the assumption of normality graphically.
+
+
+KO_weights = mice_data$weight[ mice_data$genotype=="KO"]
+WT_weights = mice_data$weight[ mice_data$genotype=="WT"]
+
+par(mfrow=c(2,2))
+
+hist(KO_weights, freq=FALSE,
+     xlab="KO", main="weights of KO mice")
+qqnorm(KO_weights)
+qqline(KO_weights)
+
+hist(WT_weights, freq=FALSE,
+     xlab="WT", main="weights of WT mice")
+qqnorm(WT_weights)
+qqline(WT_weights)
+
+par(mfrow=c(1,1)) # removes setting in par
+
+
+# KO mice: normality assumption seems ok
+
+# WT mice:
+# histogram - suggests no normality (but small sample size for this)
+# QQ Plot - quite ok
+
+
+
+# Welch's t-test is fairly robust if the deviation from normality is not extreme,
+# so let's continue
+
+# 2) Make an appropriate plot to visualize the mouse weights grouped by genotype.
+
+boxplot(weight ~ genotype, data=mice_data , 
+        col=c('orange','blue'),
+        main="mice weight by genotype")
+
+# 3) Perform a test to see whether the mouse weight is different between the two genotypes.
+t.test(KO_weights, WT_weights) #NB : here the data is not paired : WT and KO mice are different individuals
+# we do not reject HO : the is no significant difference between WT and KO mice mean weights for any reasonable significance level
+# (p=0.8514)
+
+
+# Alternative: Use non-parametric test
+# Does not assume normally distributed data
+# However, assumes the same standard deviation in all groups (while Welch's t test does not)
+# Therefore, probably not better
+
+wilcox.test(weight ~ genotype, data=mice_data)
+# no significative difference between the mean weights of WT and KO mice
+
+
+
+
+# 4) Repeat step 1 to 3 for the diet variable.
+
+HFD_weights = mice_data$weight[mice_data$diet=="HFD"]
+CHOW_weights = mice_data$weight[mice_data$diet=="CHOW"]
+
+par(mfrow=c(2,2))
+
+hist(HFD_weights, freq=FALSE,
+     xlab="HFD", main="weights of HFD mice")
+qqnorm(HFD_weights)
+qqline(HFD_weights)
+
+hist(CHOW_weights, freq=FALSE,
+     xlab="CHOW", main="weights of CHOW mice")
+qqnorm(CHOW_weights)
+qqline(CHOW_weights)
+
+par(mfrow=c(1,1))
+# all seems OK
+
+boxplot(weight ~ diet, data=mice_data, col=c('orange','blue'), main="Mouse Weight by Diet")
+
+t.test(HFD_weights, CHOW_weights ) #NB : here the data is not paired : HFD and CHOW mice are different individuals
+# we can reject HO : there is a significant difference between HFD and CHOW mice mean weights at significance level p=0.05
+# (CHOW weight are on average lower)
+
+
+# Out of curiosity, let's see what result we would get with the non-parametric Whitney Mann U test (implemented in wilcox.test)
+wilcox.test(weight ~ diet, data=mice_data)
+
diff --git a/solutions/R_practice11_solution.R b/solutions/R_practice11_solution.R
new file mode 100644
index 0000000..fbbff86
--- /dev/null
+++ b/solutions/R_practice11_solution.R
@@ -0,0 +1,31 @@
+
+# First Steps in R, Practice 11
+
+# 1) Load the package MASS using library(). (You may need to install it first).
+# Load the dataset Pima.tr using data(). Use ? to get an idea which variables it contains.
+
+library(MASS)
+data(Pima.tr)
+?Pima.tr
+
+
+#2) Hypothesis: Blood glucose level (glu) is associated with diastolic blood pressure (bp).
+#   Run a linear model to test the hypothesis.
+
+model_fit <- lm(bp~glu, data=Pima.tr)
+summary(model_fit)
+
+
+# the linear association between glucose and blood pressure is significant at the 0.05 alpha level
+# (p = 0.000115)
+
+# 3) Visualize the fit with a scatter plot and a trend line.
+par(mfrow=c(1,1))
+plot(Pima.tr$glu, Pima.tr$bp,  pch=20, xlab="blood glucose", ylab="blood pressure")
+abline(model_fit, col='red' , lwd=3)
+
+# 4) Check the assumption of normality of the residuals graphically.
+par(mfrow=c(2,1))
+plot(model_fit, which=1)
+plot(model_fit, which=2)
+# Looks quite ok
diff --git a/solutions/R_practice7_solution.R b/solutions/R_practice7_solution.R
index 4c5dd82..546583e 100644
--- a/solutions/R_practice7_solution.R
+++ b/solutions/R_practice7_solution.R
@@ -1,90 +1,48 @@
-# First Steps in R, Practice 7
-
-# Import the mouse data from the file mice_data_mod.csv. 
-#
-
-
-# 1) Check your data frame: did it load correctly? Make sure genotype and diet are factor variables.
-
-getwd() # check where you are. If you didn't change anything, you will be in the folder with the .Rproj file (rpoject root)
-
-mice_data <- read.table("course_datasets/mice_data_mod.csv", header=TRUE, sep=",") # define classes for columns
-
-str(mice_data)
-
-
-# 2) Convert genotype and diet to factor variables
-# define the order of factor levels
-mice_data$genotype <- factor(mice_data$genotype, levels=c("WT", "KO"))
-mice_data$diet <- factor(mice_data$diet, levels=c("CHOW", "HFD"))
-
-#
-
-# 3) Plot an histogram of mouse weight and customize it with colours, labels, title and represent the density line on top.
-hist(mice_data$weight,
-     freq=FALSE, breaks=8,
-     main="Mouse Weight",
-     col="orange" ,
-     xlab="weight")
-lines(density(mice_data$weight), col='blue')
-# Note: freq=FALSE makes the histogram density based, which makes it scale well with the density line
-
-
-# 4) Make a scatter plot of respiratory rate against mouse weights using the function plot().
-#    Function arguments:
-#     - use solid circles as plotting symbol
-#     - add a title
-#     - customise the axis labels  (“Weight [g]”, “Respiratory Rate [bpm]”)
-#     - colour the points by genotype. 
-# Add a legend for the genotype.
-
-
-plot(mice_data$weight,mice_data$respiratoryRate, 
-     pch=19,
-     main="Respiratory Rate vs Weight in Mice",
-     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
-     col=c("orange", "blue")[mice_data$genotype]
-     )
-
-legend("bottomright",
-       legend=levels(mice_data$genotype),
-       col=c("orange","blue"),
-       pch=19)
-
-abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
-       col="black", lwd=1.5)
-
-# 5) Make boxplots of weights from WT and KO mice. Customize with title, labels, colors.
-
-boxplot(weight ~ genotype, data= mice_data,
-        col=c("orange", "blue"),
-        main="Mouse Weight by Genotype"
-)
-points(weight ~ genotype, data= mice_data)
-
-
-# 6) Optional: Repeat 4 with diet instead of genotype.
-
-
-plot(mice_data$weight,mice_data$respiratoryRate, 
-     pch=19,
-     main="Respiratory Rate vs Weight in Mice",
-     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
-     col=c("darkred", "goldenrod")[mice_data$diet]
-)
-
-legend("bottomright",
-       legend=levels(mice_data$diet),
-       col=c("darkred","goldenrod"),
-       pch=19)
-
-abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
-       col="black", lwd=1.5)
-
-# 7) Optional: Repeat 5 with diet instead of genotype.
-
-boxplot(weight ~ diet, data= mice_data,
-        col=c("darkred", "goldenrod"),
-        main="Mouse Weight by Diet"
-)
-points(weight ~ diet, data= mice_data)
\ No newline at end of file
+# First Steps in R, Practice 7
+
+# Import the mouse data from the file mice_data_mod.csv. 
+#
+
+
+# 1) Check your data frame: did it load correctly? Make sure genotype and diet are factor variables.
+
+getwd() # check where you are. If you didn't change anything, you will be in the folder with the .Rproj file (rpoject root)
+
+mice_data <- read.table("course_datasets/mice_data_mod.csv", header=TRUE, sep=",") # define classes for columns
+
+str(mice_data)
+
+
+# 2) Convert genotype and diet to factor variables
+# define the order of factor levels
+mice_data$genotype <- factor(mice_data$genotype, levels=c("WT", "KO"))
+mice_data$diet <- factor(mice_data$diet, levels=c("CHOW", "HFD"))
+
+
+
+
+# 3) Make a scatter plot of respiratory rate against mouse weights using the function plot().
+#    Function arguments:
+#     - use solid circles as plotting symbol
+#     - add a title
+#     - customise the axis labels  (“Weight [g]”, “Respiratory Rate [bpm]”)
+#     - colour the points by genotype. 
+
+plot(mice_data$weight,mice_data$respiratoryRate, 
+     pch=19,
+     main="Respiratory Rate vs Weight in Mice",
+     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
+     col=c("orange", "blue")[mice_data$genotype]
+     )
+
+# 4) Fit a trend line using the function abline()
+abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
+       col="black", lwd=1.5)
+
+# 5) Add a legend for the genotype.
+legend("bottomright",
+       legend=levels(mice_data$genotype),
+       col=c("orange","blue"),
+       pch=19)
+
+
diff --git a/solutions/R_practice8_solution.R b/solutions/R_practice8_solution.R
index 70c748e..beef408 100644
--- a/solutions/R_practice8_solution.R
+++ b/solutions/R_practice8_solution.R
@@ -1,84 +1,30 @@
-# First Steps in R, Practice 8
-
-# 1) Make a multi-panel figure with the four graphics on one page, exporting the figure to a png file.
-#    Set width and height arguments in the call to png() to make it look nice.
-
-pdf("mice_data_plots_by_genotype.pdf", width=7, height=7, paper="a4") 
-
-# we want to make 4 plots on the same panel -> 2 rows and 2 columns
-par(mfrow=c(2,2))
-
-# Plot 1
-plot(mice_data$weight,mice_data$respiratoryRate, 
-     pch=19,
-     main="Respiratory Rate vs Weight in Mice",
-     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
-     col=c("orange", "blue")[mice_data$genotype]
-)
-
-legend("bottomright",
-       legend=levels(mice_data$genotype),
-       col=c("orange","blue"),
-       pch=19)
-
-abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
-       col="black", lwd=1.5)
-
-# Plot 2
-boxplot(weight ~ genotype, data= mice_data,
-        col=c("orange", "blue"),
-        main="Mouse Weight by Genotype"
-)
-points(weight ~ genotype, data= mice_data)
-
-
-# Plot 3
-plot(mice_data$weight,mice_data$respiratoryRate, 
-     pch=19,
-     main="Respiratory Rate vs Weight in Mice",
-     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
-     col=c("darkred", "goldenrod")[mice_data$diet]
-)
-
-legend("bottomright",
-       legend=levels(mice_data$diet),
-       col=c("darkred","goldenrod"),
-       pch=19)
-
-abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
-       col="black", lwd=1.5)
-
-# Plot4
-boxplot(weight ~ diet, data= mice_data,
-        col=c("darkred", "goldenrod"),
-        main="Mouse Weight by Diet"
-)
-points(weight ~ diet, data= mice_data)
-
-dev.off()
-
-
-# 2) Optional: Export the histogram (3 from previous exercise) to a png file. 
-# Set width and height arguments in the call to png() to make it look nice.
-
-# png: width and height are in pixels by default
-
-png("hist_weight.png", width=800, height=600)
-hist(mice_data$weight,
-     freq=FALSE, breaks=8,
-     main="Mouse Weight",
-     col="orange" ,
-     xlab="weight")
-lines(density(mice_data$weight), col='blue')
-dev.off()
-
-# 3) Look at the multi-panel figure. Are your impressions about mouse weight from yesterday's exploration of data summaries confirmed by today's visualizations?
-# Yes.
-# Genotpye: The two genotypes (KO v WT) have very similar mean weights.
-# Diet: Mice on HFD are heavier on average than mice on CHOW diet, and their weights are more variable.
-
-
-
-
-
-
+mice_data <- read.table("course_datasets/mice_data_mod.csv", header=TRUE, sep=",") # define classes for columns
+mice_data$genotype <- factor(mice_data$genotype, levels=c("WT", "KO"))
+mice_data$diet <- factor(mice_data$diet, levels=c("CHOW", "HFD"))
+
+
+# 1) Plot an histogram of mouse weight and customize it with colours, labels, title and represent the density line on top.
+hist(mice_data$weight,
+     freq=FALSE, breaks=8,
+     main="Mouse Weight",
+     col="orange" ,
+     xlab="weight")
+lines(density(mice_data$weight), col='blue')
+# Note: freq=FALSE makes the histogram density based, which makes it scale well with the density line
+
+# 2) Make boxplots of weights from WT and KO mice. Customize with title, labels, colors.
+
+boxplot(weight ~ genotype, data= mice_data,
+        col=c("orange", "blue"),
+        main="Mouse Weight by Genotype"
+)
+points(weight ~ genotype, data= mice_data)
+
+
+# 3) Optional: Repeat 2 with diet instead of genotype.
+
+boxplot(weight ~ diet, data= mice_data,
+        col=c("darkred", "goldenrod"),
+        main="Mouse Weight by Diet"
+)
+points(weight ~ diet, data= mice_data)
\ No newline at end of file
diff --git a/solutions/R_practice9_solution.R b/solutions/R_practice9_solution.R
index 69cc540..30f0b53 100644
--- a/solutions/R_practice9_solution.R
+++ b/solutions/R_practice9_solution.R
@@ -1,95 +1,84 @@
-# First Steps in R, Practice 9
-
-# Come back to the mice data-set stored in the "mice_data" data frame.
-
-mice_data <- read.table( "course_datasets/mice_data.csv" , header=T , sep=",",
-                         colClasses=c("factor", "factor", "numeric")) # define classes for columns  )
-
-mice_data$genotype <- factor(mice_data$genotype, levels=c("WT", "KO"))
-
-
-# 1)  Considering WT mice weight and KO mice weight separately, check the assumption of normality graphically.
-
-
-KO_weights = mice_data$weight[ mice_data$genotype=="KO"]
-WT_weights = mice_data$weight[ mice_data$genotype=="WT"]
-
-par(mfrow=c(2,2))
-
-hist(KO_weights, freq=FALSE,
-     xlab="KO", main="weights of KO mice")
-qqnorm(KO_weights)
-qqline(KO_weights)
-
-hist(WT_weights, freq=FALSE,
-     xlab="WT", main="weights of WT mice")
-qqnorm(WT_weights)
-qqline(WT_weights)
-
-par(mfrow=c(1,1)) # removes setting in par
-
-
-# KO mice: normality assumption seems ok
-
-# WT mice:
-# histogram - suggests no normality (but small sample size for this)
-# QQ Plot - quite ok
-
-
-
-# Welch's t-test is fairly robust if the deviation from normality is not extreme,
-# so let's continue
-
-# 2) Make an appropriate plot to visualize the mouse weights grouped by genotype.
-
-boxplot(weight ~ genotype, data=mice_data , 
-        col=c('orange','blue'),
-        main="mice weight by genotype")
-
-# 3) Perform a test to see whether the mouse weight is different between the two genotypes.
-t.test(KO_weights, WT_weights) #NB : here the data is not paired : WT and KO mice are different individuals
-# we do not reject HO : the is no significant difference between WT and KO mice mean weights for any reasonable significance level
-# (p=0.8514)
-
-
-# Alternative: Use non-parametric test
-# Does not assume normally distributed data
-# However, assumes the same standard deviation in all groups (while Welch's t test does not)
-# Therefore, probably not better
-
-wilcox.test(weight ~ genotype, data=mice_data)
-# no significative difference between the mean weights of WT and KO mice
-
-
-
-
-# 4) Repeat step 1 to 3 for the diet variable.
-
-HFD_weights = mice_data$weight[mice_data$diet=="HFD"]
-CHOW_weights = mice_data$weight[mice_data$diet=="CHOW"]
-
-par(mfrow=c(2,2))
-
-hist(HFD_weights, freq=FALSE,
-     xlab="HFD", main="weights of HFD mice")
-qqnorm(HFD_weights)
-qqline(HFD_weights)
-
-hist(CHOW_weights, freq=FALSE,
-     xlab="CHOW", main="weights of CHOW mice")
-qqnorm(CHOW_weights)
-qqline(CHOW_weights)
-
-par(mfrow=c(1,1))
-# all seems OK
-
-boxplot(weight ~ diet, data=mice_data, col=c('orange','blue'), main="Mouse Weight by Diet")
-
-t.test(HFD_weights, CHOW_weights ) #NB : here the data is not paired : HFD and CHOW mice are different individuals
-# we can reject HO : there is a significant difference between HFD and CHOW mice mean weights at significance level p=0.05
-# (CHOW weight are on average lower)
-
-
-# Out of curiosity, let's see what result we would get with the non-parametric Whitney Mann U test (implemented in wilcox.test)
-wilcox.test(weight ~ diet, data=mice_data)
-
+# First Steps in R, Practice 9
+
+# 1) Make a multi-panel figure with the four graphics on one page, exporting the figure to a png file.
+#    Set width and height arguments in the call to png() to make it look nice.
+
+pdf("mice_data_plots_by_genotype.pdf", width=7, height=7, paper="a4") 
+
+# we want to make 4 plots on the same panel -> 2 rows and 2 columns
+par(mfrow=c(2,2))
+
+# Plot 1
+plot(mice_data$weight,mice_data$respiratoryRate, 
+     pch=19,
+     main="Respiratory Rate vs Weight in Mice",
+     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
+     col=c("orange", "blue")[mice_data$genotype]
+)
+
+legend("bottomright",
+       legend=levels(mice_data$genotype),
+       col=c("orange","blue"),
+       pch=19)
+
+abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
+       col="black", lwd=1.5)
+
+# Plot 2
+boxplot(weight ~ genotype, data= mice_data,
+        col=c("orange", "blue"),
+        main="Mouse Weight by Genotype"
+)
+points(weight ~ genotype, data= mice_data)
+
+
+# Plot 3
+plot(mice_data$weight,mice_data$respiratoryRate, 
+     pch=19,
+     main="Respiratory Rate vs Weight in Mice",
+     xlab="Weight [g]", ylab="Respiratory Rate [bpm]",
+     col=c("darkred", "goldenrod")[mice_data$diet]
+)
+
+legend("bottomright",
+       legend=levels(mice_data$diet),
+       col=c("darkred","goldenrod"),
+       pch=19)
+
+abline(lm(mice_data$respiratoryRate ~ mice_data$weight),
+       col="black", lwd=1.5)
+
+# Plot4
+boxplot(weight ~ diet, data= mice_data,
+        col=c("darkred", "goldenrod"),
+        main="Mouse Weight by Diet"
+)
+points(weight ~ diet, data= mice_data)
+
+dev.off()
+
+
+# 2) Optional: Export the histogram (3 from previous exercise) to a png file. 
+# Set width and height arguments in the call to png() to make it look nice.
+
+# png: width and height are in pixels by default
+
+png("hist_weight.png", width=800, height=600)
+hist(mice_data$weight,
+     freq=FALSE, breaks=8,
+     main="Mouse Weight",
+     col="orange" ,
+     xlab="weight")
+lines(density(mice_data$weight), col='blue')
+dev.off()
+
+# 3) Look at the multi-panel figure. Are your impressions about mouse weight from yesterday's exploration of data summaries confirmed by today's visualizations?
+# Yes.
+# Genotpye: The two genotypes (KO v WT) have very similar mean weights.
+# Diet: Mice on HFD are heavier on average than mice on CHOW diet, and their weights are more variable.
+
+
+
+
+
+