10-Data-Engineering-2.Rmd

---
title: "Veri Mühendisliği-2 (Data Engineering -2)" by Tirendaz Akademi
---
#Verileri Gruplama

```{r}
head(iris)
```
```{r}
iris_summary<-aggregate(cbind(Petal.Length, Petal.Width) ~ Species, 
                        data=iris, FUN=mean)
iris_summary
```
```{r}
library("data.table")
iris_dt<-as.data.table(iris)
iris_dt<-iris_dt[, .(Petal.Length=mean(Petal.Length),
                     Petal.Width=mean(Petal.Width)), 
                 by=.(Species)]
iris_dt
```
```{r}
library("dplyr")
iris_dplyr<-iris %>% 
  group_by(.,Species) %>% 
  summarize(.,Petal.Length=mean(Petal.Length),
            Petal.Width=mean(Petal.Width)) %>% 
  ungroup(.)
iris_dplyr
```
#Çoklu Veri Dönüşümü

```{r}
ID<-1:5
Mat<-sample(100,5)
Sta<-sample(100,5)
A<-data.frame(ID,Mat,Sta)
A
```

```{r}
Eng<-sample(100,5)
B<-data.frame(ID,Mat,Eng)
B
```

```{r}
ID<-6:10
Mat<-sample(100,5)
Sta<-sample(100,5)
C<-data.frame(ID,Mat,Sta)
C
```
```{r}
AC<-rbind(A,C)
AC
```
```{r}
bind_rows(list(A,C))
```
```{r}
A$class<-c("A")
C$class<-c("C")
AC<-rbind(A,C)
AC
```
```{r}
t<-split(AC, AC$class)
t
```
```{r}
dt<-as.data.table(AC)
f<-function(.BY,.SD){max(.SD$Mat)}
dt[, max_mat :=f(.BY,.SD), by=class]
dt
```
```{r}
AC %>% 
  group_by(.,class) %>% 
  mutate(.,max_mat=max(Mat)) %>% 
  ungroup(.)
```
```{r}
cbind(A,B)
```

```{r}
cbind(as.data.table(A),as.data.table(C))
```
```{r}
bind_cols(list(A,C))
```

```{r}
merge(A,B,by="ID")
```
```{r}
ID<-c("a","b","c","d")
CS<-c(85,90,95,80)
D<-data.frame(ID,CS)

ID<-c("a","b","c","e")
CS<-c(80,90,90,70)
E<-data.frame(ID,CS)

(D)
(E)
```
```{r}
merge(D,E, by="ID", all.x = TRUE)
```
```{r}
D_dt<-as.data.table(D)
E_dt<-as.data.table(E)
E_dt[D_dt, on="ID"]
```
```{r}
left_join(D,E,by="ID")
```
```{r}
merge(D,E, by="ID")
```

```{r}
merge(D_dt,E_dt, by="ID")
```
```{r}
inner_join(D,E,by="ID")
```
```{r}
merge(D,E, by="ID", all=TRUE)
```
```{r}
merge(D_dt,E_dt,by="ID", all=TRUE)
```
#Dönüşümleri Yeniden Şekillerdirme

```{r}
library("datasets")
df<-Seatbelts
head(df)
```

```{r}
str(df)
```
```{r}
library("xts")
dates<-index(as.xts(time(df)))
df<-data.frame(df)
df$dates<-dates
head(df)
```
```{r}
head(df$dates)
tail(df$dates)
```
```{r}
df<-df[(df$dates>=as.yearmon("Oca 1982")) & 
        (df$dates<=as.yearmon("Ara 1983")), , drop=FALSE ]
str(df)
```
```{r}
df$dates<-as.Date(df$dates)
df$law<-ifelse(df$law==1, "new law", "pre law")
df<-df[,c("dates","DriversKilled","front","rear","law")]
head(df)
```
```{r}
library("ggplot2")
ggplot(df, aes(x=dates, y=DriversKilled, color =law, shape=law))+
  geom_point()+
  geom_smooth(se=FALSE)+
  ggtitle("Aylık Sürücü Ölümleri")
```
```{r}
df_long<-melt.data.table(as.data.table(df), id.vars = NULL,
                         measure.vars = c("DriversKilled","front","rear"),
                         variable.name = "victim_type",
                         value.name = "nvictims")
head(df_long)
```
```{r}
library("cdata")
df_long2<-unpivot_to_blocks(
  df,
  nameForNewKeyColumn = "victim_type",
  nameForNewValueColumn = "nvictims",
  columnsToTakeFrom = c("DriversKilled","front","rear")
)
head(df_long2)
```
```{r}
library("tidyr")
df_long3<-gather(
  df,
  key=victim_type,
  value=nvictims,
  DriversKilled, front,rear
)
head(df_long3)
```

```{r}
ggplot(df_long3, aes(x=dates, y=nvictims, color =law, shape=law))+
  geom_point()+
  geom_smooth(se=FALSE)+
  facet_wrap(~victim_type, ncol=1,scale="free_y")+
  ggtitle("Oturma Pozisyonuna Göre Ölümleri")
```
```{r}
CH<-data.frame(ChickWeight)
str(CH)
```
```{r}
CH$Diet<-NULL
head(CH)
```

```{r}
CH_wide<-dcast.data.table(as.data.table(CH), Chick~Time,
                          value.var = "weight")
head(CH_wide)
```
```{r}
CH_wide2<-pivot_to_rowrecs(
  CH,
  columnToTakeKeysFrom = "Time",
  columnToTakeValuesFrom = "weight",
  rowKeyColumns = "Chick"
  )
head(CH_wide2)
```

```{r}
CH_wide3<-spread(CH,
                key=Time,
                value = weight)
head(CH_wide3)
```