Transaksi mie instant
meningkat menjelang tanggal tua.
 |
| Tampilan dashboard shiny |
Teringat masa kuliah, ketika menjelang tanggal tua uang saku
menipis kira kira menu apa yang akan kita santap untuk mengisi perut. Sudah
menjadi rahasia umum bagi mahasiswa perantauan saat itu, ya apalagi kalau bukan
mie instant. Kebiasaan ini ternyata masih berlanjut, padahal kebanyakan makan
mi instan itu kan tidak sehat. Simpel
saja sih alasannya daripada perut lapar kan tidak bisa belajar.
Fenomena tersebut yang mengusik pikiran saya untuk
meyakinkan apakah benar pada tanggal tua transaksi mie instan meningkat. Untuk
menjawab hal tersebut saya mencoba
menganalisa data transaksi dari sebuah minimarket di sekitar kampus.
 |
| Gb1 |
Karena datanya tidak besar, saya gunakan pengamatan empiris
lewat grafik, yang diproses dengan R programming, divisualkan dengan shiny aplikasi, berikut script R nya:
 |
| Gb2 |
#--- Persiapan
Data
head(dm)
year month date_of_month day_of_week
mie_instan
1 2000 1 1 6 120
2 2000 1 2 7 99
3 2000 1 3 1 123
4 2000 1 4 2 132
5 2000 1 5 3 133
6 2000 1 6 4 100
dim(dm)
[1] 5479 5
library(shiny) #untuk
visualisasi
library(ggplot2) #untuk Grafis
library(dplyr) #Untuk proses
#---Panel UI
Shiny Application
ui <- fluidPage(
# Some custom CSS for a smaller font for
preformatted text
tags$head(
tags$style(HTML("
pre, table.table {
font-size: smaller;
}
")
),
fluidRow(
column(width = 4, wellPanel(
radioButtons("plot_type",
"Plot type",
c("P1-Sales-2000",
"P2-Sales-2001","P3-Sales-2002","P4-Sales-2003","P5-Sales-2004",
"P6-Sales-2005","P7-Sales-2006","P8-Sales-2007",
"P9-Sales-2008","P10-Sales-2009","P11-Sales-2010",
"P12-Sales-2011","P13-Sales-2012","P14-Sales-2013","P15-Sales-2014",
"Sales-Total-Year","Salest-Total-Split-Year")
)
)),
column(width = 4,
# In a plotOutput, passing values
for click, dblclick, hover, or brush
# will enable those interactions.
plotOutput("plot1", brush
= "plot_brush", width = 500, height = 350),
verbatimTextOutput("info"))
)
)
)
#---Data
proccess
library(shiny)
library(ggplot2)
library(dplyr)
#---Data local
dm <-
read.csv("G:/new-program/mie instant/rsconnect/Data/mi.csv")
#---count number days each
year
#n_day_year1 <- dm %>% count(year) %>% arrange(desc(year))
%>% slice(1:15)
#n_day_year1
# A tibble: 15 ? 2
# year
n
# <int> <int>
#1 2014
365
#2 2013
365
#3 2012
366
#4 2011
365
#5 2010
365
#6 2009
365
#7 2008
366
#8 2007
365
#9 2006
365
#10 2005
365
#11 2004
366
#12 2003
365
#13 2002
365
#14 2001
365
#15 2000
366
#---Mean per year
d1 <- dm[1:366,]
gd1 <- d1 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P1 <- ggplot(gd1, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2000",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d2 <- dm[367:731,]
gd2 <- d2 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P2 <- ggplot(gd2, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2001",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d3 <- dm[732:1096,]
gd3 <- d3 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P3 <- ggplot(gd3, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2002",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d4 <- dm[1097:1461,]
gd4 <- d4 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P4 <- ggplot(gd4, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2003",subtitle = "Crafted by www.realrapplication.blogspot.com")
+
labs(x = "Day", y =
"Mean") +
theme_bw()
d5 <- dm[1462:1821,]
gd5 <- d5 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P5 <- ggplot(gd5, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2004",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d6 <- dm[1828:2192,]
gd6 <- d6 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P6 <- ggplot(gd6, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2005",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d7 <- dm[2193:2557,]
gd7 <- d7 %>%
group_by(date_of_month)%>%
mutate(md=mean(mie_instan))
P7 <- ggplot(gd7, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2006",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d8 <- dm[2558:2922,]
gd8 <- d8 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P8 <- ggplot(gd8, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2007",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d9 <- dm[2923:3288,]
gd9 <- d9 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P9 <- ggplot(gd9, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2008",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d10 <- dm[3289:3653,]
gd10 <- d10 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P10 <- ggplot(gd10, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2009",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d11 <- dm[3654:4018,]
gd11 <- d11 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P11 <- ggplot(gd11, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2010",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d12 <- dm[4019:4383,]
gd12 <- d12 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P12 <- ggplot(gd12, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2011",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d13 <- dm[4384:4749,]
gd13 <- d13 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P13 <- ggplot(gd13, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2012",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d14 <- dm[4750:5114,]
gd14 <- d14 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P14 <- ggplot(gd14, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2013",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
d15 <- dm[5115:5479,]
gd15 <- d15 %>%
group_by(date_of_month) %>%
mutate(md=mean(mie_instan))
P15 <- ggplot(gd15, aes(x =
date_of_month, y = md)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
ggtitle("Sales Mie-Instant
2014",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
#---Binding mean per year
Py <-
rbind(gd1,gd2,gd3,gd4,gd5,gd6,gd7,gd8,gd9,
gd10,gd11,gd12,gd13,gd14,gd15)
#---Grafik mean th 2000-2014
Sales_Total_Year <-
ggplot(Py, aes(x = date_of_month, y = md)) +
geom_line() +
ggtitle("Sales Mie Instant per month
2000-2014",subtitle = "Crafted by
www.realrapplication.blogspot.com") +
labs(x = "Day", y =
"Mean") +
theme_bw()
#---Grafik split per th
Salest_Total_Split_Year <-
ggplot(Py,aes(x = date_of_month, y = md, group = year)) +
geom_line(color = "dark red", size
= 1, na.rm = T) +
facet_wrap(~ year) +
labs(title = "Sales Mie-instant per
month per year\n Crafted by
www.realrapplication.blogspot.com") +
labs(subtitle = "2000-14") +
labs(x = "Day", y =
"Mean") +
theme_bw()
#---Panel
server Shiny Application
server <- function(input,
output) {
output$plot1 <- renderPlot({
if (input$plot_type ==
"P1-Sales-2000") {
P1
} else if (input$plot_type ==
"P2-Sales-2001") {
P2
} else if (input$plot_type ==
"P3-Sales-2002") {
P3
}else if (input$plot_type ==
"P4-Sales-2003") {
P4
}else if (input$plot_type ==
"P5-Sales-2004") {
P5
}else if (input$plot_type ==
"P6-Sales-2005") {
P6
}else if (input$plot_type ==
"P7-Sales-2006") {
P7
}else if (input$plot_type ==
"P8-Sales-2007") {
P8
}else if (input$plot_type ==
"P9-Sales-2008") {
P9
}else if (input$plot_type ==
"P10-Sales-2009") {
P10
}else if (input$plot_type ==
"P11-Sales-2010") {
P11
}else if (input$plot_type ==
"P12-Sales-2011") {
P12
}else if (input$plot_type ==
"P13-Sales-2012") {
P13
}else if (input$plot_type ==
"P14-Sales-2013") {
P14
}else if (input$plot_type ==
"P15-Sales-2014") {
P15
}else if (input$plot_type ==
"Sales-Total-Year") {
Sales_Total_Year
}else if (input$plot_type ==
"Salest-Total-Split-Year") {
Salest_Total_Split_Year
}
})
output$info <- renderPrint({
brushedPoints(Py, input$plot_brush)
})
}
shinyApp(ui=ui,server=server)
Kesimpulan:
1. Mengamati
grafik P1-P15, yaitu penjualan mie instan per bulan, nampak jelas penjualan
meningkat di sekitar tgl tua tiap bulan. Berdasar data empiris transaksi per
bulan tsb makin jelas divisualkan dalam
Grafik "Sales Mie Instant per month 2000-2014".Lihat Gb1
2. Kalo diamati
lebih dalam juga terjadi fenomena transaksi mie instant periodically per 7
hari, analisa lebih detail bisa dilanjutkan lain waktu.
3. Bagi pemilik
minimarket ini adalah info yang sangat penting, karena berhubungan dengan
persediaan barang.Lihat Gb2
4. Dengan R
programming prosesnya cepat, pengalaman penulis pernah memproses data transaksi
terdiri dari 1/2 juta baris, tidak terlalu lama, running di laptop Azus 1.65
HZ, RAM 4 GB.
Referensi:
Tidak ada komentar:
Posting Komentar