#contoh soal 1

Data=read.table(file.choose(), header = T)
Data
##               Provinsi   IPM
## 1                 ACEH 74.03
## 2       SUMATERA_UTARA 74.02
## 3       SUMATERA_BARAT 74.49
## 4                 RIAU 74.79
## 5                JAMBI 73.43
## 6     SUMATERA_SELATAN 72.30
## 7             BENGKULU 73.39
## 8              LAMPUNG 71.81
## 9  KEP_BANGKA_BELITUNG 73.33
## 10            KEP_RIAU 77.97
## 11         DKI_JAKARTA 83.08
## 12          JAWA_BARAT 74.43
## 13         JAWA_TENGAH 73.88
## 14       DI_YOGYAKARTA 81.55
## 15          JAWA_TIMUR 74.09
## 16              BANTEN 74.48
## 17                BALI 77.76
## 18 NUSA_TENGGARA_BARAT 70.93
## 19 NUSA_TENGGARA_TIMUR 67.39
## 20    KALIMANTAN_BARAT 70.13
## 21   KALIMANTAN_TENGAH 72.73
## 22  KALIMANTAN_SELATAN 73.03
## 23    KALIMANTAN_TIMUR 78.83
## 24    KALIMANTAN_UTARA 73.02
## 25      SULAWESI_UTARA 75.03
## 26     SULAWESI_TENGAH 71.56
## 27    SULAWESI_SELATAN 74.05
## 28   SULAWESI_TENGGARA 73.48
## 29           GORONTALO 71.23
## 30      SULAWESI_BARAT 68.20
## 31              MALUKU 71.57
## 32        MALUKU_UTARA 71.03
## 33         PAPUA_BARAT 67.02
## 34    PAPUA_BARAT_DAYA 68.63
## 35               PAPUA 73.00
## 36       PAPUA_SELATAN 67.90
## 37        PAPUA_TENGAH 59.75
## 38    PAPUA_PEGUNUNGAN 53.42
length(Data$IPM)
## [1] 38
mean(Data$IPM)
## [1] 72.38842
median(Data$IPM)
## [1] 73.18
range(Data$IPM)
## [1] 53.42 83.08
max(Data$IPM)
## [1] 83.08
min(Data$IPM)
## [1] 53.42
var(Data$IPM)
## [1] 26.52484
sd(Data$IPM)
## [1] 5.150227
quantile(Data$IPm)
##   0%  25%  50%  75% 100% 
##   NA   NA   NA   NA   NA

#contoh soal 2

library(ggplot2)
ggplot(Data, aes(x=IPM)) + 
  geom_histogram(
    binwidth = 5,
    boundary = 60,
    fill = "pink",
    color = "black"
  ) +
scale_x_continuous(
  breaks = seq(60,85,by=5),
  limits=c(60,85)
) +
  labs(
    title = "IPM di Indonesia",
    x = "IPM",
    y = "Freq"
  ) +
  theme_classic() +
  theme (
    axis.text.x = element_text (angle = 0, hjust = 0.5),
    plot.title = element_text(
      hjust = 0.5,
      face = "bold"
    )
  )
## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).

ggplot(Data, aes(x = "", y = IPM)) +
  geom_boxplot(fill = "red") +
  labs(title = "Box Plot IPM INdonesia 2024", x = "", y = "IPM") +
  theme_minimal()+
  theme (plot.title = element_text (hjust = 0.5, face = "bold"))

library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
library(tidyverse)
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ forcats   1.0.1     ✔ stringr   1.6.0
## ✔ lubridate 1.9.5     ✔ tibble    3.3.1
## ✔ purrr     1.2.2     ✔ tidyr     1.3.2
## ✔ readr     2.2.0
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
library(ggplot2)

Data_Kalimantan <- Data %>%
  filter(str_starts(Provinsi, "KALIMANTAN"))

Data_Kalimantan
##             Provinsi   IPM
## 1   KALIMANTAN_BARAT 70.13
## 2  KALIMANTAN_TENGAH 72.73
## 3 KALIMANTAN_SELATAN 73.03
## 4   KALIMANTAN_TIMUR 78.83
## 5   KALIMANTAN_UTARA 73.02
ggplot(Data_Kalimantan, aes(x = Provinsi, y = IPM, fill = Provinsi)) +
  geom_col() +
  labs(
    title = "IPM Pulau Kalimantan Tahun 2024",
    x = "Provinsi",
    y = "IPM"
  ) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    plot.title = element_text(hjust = 0.5, face = "bold"),
    legend.position = "none"
  )

# Load library
library(ggplot2)
# Baca data
data = read.table(file.choose(), header = T, sep = "\t")
data
##         Nama_Produk  Kategori Penjualan
## 1             Beras   Sembako        35
## 2     Minyak_Goreng   Sembako        75
## 3        Gula_Pasir   Sembako        60
## 4        Mie_Instan   Makanan       150
## 5           Biskuit   Makanan        85
## 6          Susu_UHT   Makanan        70
## 7         The_Celup   Makanan        55
## 8       Sabun_Mandi Perawatan        90
## 9             Sampo Perawatan        65
## 10       Pasta_Gigi Perawatan        70
## 11 Sabun_Cuci_ Muka Perawatan        45
# Buat barplot dengan ggplot2
ggplot(data, aes(x = Nama_Produk, y = Penjualan, fill = Kategori)) +
  geom_bar(stat = "identity") +
  labs(
    title = "Jumlah Penjualan Produk Toko MUTIA",
    x = "Produk",
    y = "Jumlah Penjualan"
  ) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    plot.title = element_text(hjust = 0.5, face = "bold")
  )

# Agregasi data (total penjualan per kategori)
total_penjualan <- aggregate(Penjualan ~ Kategori, data, sum)
total_penjualan
##    Kategori Penjualan
## 1   Makanan       360
## 2 Perawatan       270
## 3   Sembako       170
# Buat barplot dengan ggplot2
ggplot(total_penjualan, aes(x = Kategori, y = Penjualan, fill = Kategori)) +
  geom_bar(stat = "identity") +
  labs(
    title = "Jumlah Penjualan per Kategori",
    x = "Kategori",
    y = "Jumlah Penjualan"
  ) +
  theme_minimal() +
  theme(
    axis.text.x = element_text(angle = 45, hjust = 1),
    plot.title = element_text(hjust = 0.5, face = "bold")
  )