> df=read.csv("titanic.csv")
> str(df)
'data.frame':	887 obs. of  6 variables:
 $ Survived: logi  FALSE TRUE TRUE TRUE FALSE FALSE ...
 $ Pclass  : int  3 1 3 1 3 3 1 3 3 2 ...
 $ Name    : Factor w/ 887 levels "Capt. Edward Gifford Crosby",..: 602 823 172 814 733 464 700 33 842 839 ...
 $ Sex     : Factor w/ 2 levels "female","male": 2 1 1 1 2 2 2 2 1 1 ...
 $ Age     : num  22 38 26 35 35 27 54 2 27 14 ...
 $ Fare    : num  7.25 71.28 7.92 53.1 8.05 ...
mean(df$Survived[df$Sex=="female"])
[1] 0.7420382
> (91+70+72)/(91+70+72+3+6+72)
[1] 0.7420382

mean(df$Survived[df$Sex=="female" & df$Pclass==1])
[1] 0.9680851
> (91)/(91+3)
[1] 0.9680851

mean(df$Survived[df$Sex=="female" & df$Pclass==2])
[1] 0.9210526
> 70/(70+6)
[1] 0.9210526

mean(df$Survived[df$Sex=="female" & df$Pclass==3])
[1] 0.5
> 72/(72+72)
[1] 0.5

> mean(df$Survived[df$Sex=="male"])
[1] 0.1902269
> mean(df$Survived[df$Sex=="male" & df$Pclass==1])
[1] 0.3688525
> mean(df$Survived[df$Sex=="male" & df$Pclass==2])
[1] 0.1574074
> mean(df$Survived[df$Sex=="male" & df$Pclass==3])
[1] 0.1370262
> (45+17+47)/(45+17+47+77+91+296)
[1] 0.1902269
> (45)/(45+77)
[1] 0.3688525
> (17)/(17+91)
[1] 0.1574074
> (47)/(47+296)
[1] 0.1370262

(91+70+72)/(91+70+72+3+6+72)=
91/(91+3)*(91+3)/(91+70+72+3+6+72)
70/(70+6)*(70+6)/(91+70+72+3+6+72)
72/(72+72)*(72+72)/(91+70+72+3+6+72)

(91+45)/(91+70+72+45+17+47) * (91+70+72+45+17+47)/877
(91+45)/(94+144) * (94+144)/877

(91+70+72)/877 .ne.
342/877 * 314/877

> a=fisher.test(fout)
> a

	Fisher's Exact Test for Count Data

data:  fout
p-value < 2.2e-16
alternative hypothesis: two.sided

> b=chisq.test(mout)
> b

	Pearson's Chi-squared test

data:  mout
X-squared = 32.328, df = 2, p-value = 9.552e-08


> b$expected
        Pclass
Survived        1       2         3
   FALSE 98.79232 87.4555 277.75218
   TRUE  23.20768 20.5445  65.24782

main difference is in 1st class survival: less was expected

young females & 24-32 males more common
> pM=hist(df$Age[df$Sex=="male"],xlim=c(0,80),breaks=seq(0,80,8),freq=F)
> pF=hist(df$Age[df$Sex=="female"],xlim=c(0,80),breaks=seq(0,80,8),freq=F)

1st class:
8-24 females, and generally older men likely
> pM=hist(df$Age[df$Sex=="male" & df$Pclass==1],xlim=c(0,80),breaks=seq(0,80,8),freq=F)
> pF=hist(df$Age[df$Sex=="female" & df$Pclass==1],xlim=c(0,80),breaks=seq(0,80,8),freq=F)

1/3 men:
young 3rd class & older 1st class men likely
> pM=hist(df$Age[df$Sex=="male" & df$Pclass==1],xlim=c(0,80),breaks=seq(0,80,8),freq=F)
> pF=hist(df$Age[df$Sex=="male" & df$Pclass==3],xlim=c(0,80),breaks=seq(0,80,8),freq=F)

survive/dead
0-8 survived; 18-24 dead
> pM=hist(df$Age[df$Survived],xlim=c(0,80),breaks=seq(0,80,8),freq=F)
> pF=hist(df$Age[! df$Survived],xlim=c(0,80),breaks=seq(0,80,8),freq=F)

count men surived/total:
0-8 best bet to survive
> pM=hist(df$Age[df$Sex=="male"],xlim=c(0,80),breaks=seq(0,80,8))
> pM=hist(df$Age[df$Sex=="male" & df$Survived],xlim=c(0,80),breaks=seq(0,80,8))

count men survive/total 1st class
very young 100%, middle age ~50%
> pF=hist(df$Age[df$Sex=="male" & df$Pclass==1],xlim=c(0,80),breaks=seq(0,80,8)) 
> pM=hist(df$Age[df$Sex=="male" & df$Survived & df$Pclass==1],xlim=c(0,80),breaks=seq(0,80,8))

count men survive/total 2nd class
very young survived
> pF=hist(df$Age[df$Sex=="male" & df$Pclass==2],xlim=c(0,80),breaks=seq(0,80,8)) 
> pM=hist(df$Age[df$Sex=="male" & df$Survived & df$Pclass==2],xlim=c(0,80),breaks=seq(0,80,8))

count men survive/total 3rd class
not good for any, but young better than old
> pF=hist(df$Age[df$Sex=="male" & df$Pclass==3],xlim=c(0,80),breaks=seq(0,80,8)) 
> pM=hist(df$Age[df$Sex=="male" & df$Survived & df$Pclass==3],xlim=c(0,80),breaks=seq(0,80,8))

count female survived
pretty good for all age groups, old was best
> pF=hist(df$Age[df$Sex=="female"],xlim=c(0,80),breaks=seq(0,80,8))
> pM=hist(df$Age[df$Sex=="female" & df$Survived ],xlim=c(0,80),breaks=seq(0,80,8))

t.test(df$Age[df$Sex=="female" & df$Pclass==1],df$Age[df$Sex=="male" & df$Pclass==1])
t = -3.3197, df = 206.89, p-value = 0.001065
females 35 vs 41

> t.test(df$Age[df$Sex=="male" & df$Pclass==3],df$Age[df$Sex=="male" & df$Pclass==1])
t = -10.402, df = 183.16, p-value < 2.2e-16
3rd: 26.5 vs 1st 41.5

> df2=df[df$Fare>0,]
> boxplot(df2$Fare ~ df2$Pclass,log="Y")
Error in plot.window(xlim = xlim, ylim = ylim, log = log, yaxs = pars$yaxs) : 
  invalid "log=Y" specification
> boxplot(df2$Fare ~ df2$Pclass,log="y")

> which.min(df2$Fare[df2$Pclass==1])
[1] 208
> (df2[df2$Pclass==1,])[208,] ...........note: ()[208,]
    Survived Pclass                    Name  Sex Age Fare
869    FALSE      1 Mr. Frans Olof Carlsson male  33    5