how to vectorise my code in r using for loop? - r

Question

I am trying to run a for loop for the below scenario
I have 100k Unique customers in my transactional table which are going to n number of stores each customer.
I am trying to loop through each customer and seeing in which unique store they are going and if new store has opened within 20 kms range to that store then he/she will go to that store and I will make their Value 1 in new data frame created.
I am initializing my code in first for loop and then repeating the same for rest of my data frame.
My code is extremely slow with for loop. I don't know how to vectorise my code. Below is the snapshot of my code. Please guide me how to make this code faster and efficient.
cust_id = c(unique(kk$Customer_ID))
i = cust_id[1]
# for initializing
s = c(0,0,0,0)
df_temp = kk[kk$Customer_ID == i]
store = c(unique(df_temp$Store_Code))
system.time(
for (j in store){
if(df_temp[Store_Code == j]$dist.km298 < 20) {
s[1] <- 1
}
if(df_temp[Store_Code == j]$dist.km299 < 20) {
s[2] <- 1
}
if(df_temp[Store_Code == j]$dist.km300 < 20) {
s[3] <- 1
}
if(df_temp[Store_Code == j]$dist.km301 < 20) {
s[4] <- 1
}
}
)
vishal <- data.table("Customer_ID" = c(i,i,i,i) , "Store_Code" =
c(60298,60299,60300,60301), "Prediction" = s)
cust_id <- cust_id[!cust_id %in% c(cust_id[1])]
# loop for all customers
count = 1
system.time(for (k in 1:length(cust_id)){
i <- cust_id[k]
# count <- count+1
# if (count == 5) {
# break
#}
s = c(0,0,0,0)
df_temp = kk[kk$Customer_ID == i]
store = c(unique(df_temp$Store_Code))
for (j in store){
#if(df_temp$Store_Code == j & df_temp$Purchase_2016 != 0 &
df_temp$Purchase_2017 == 0){
if(df_temp[Store_Code == j]$dist.km298 < 20) {
s[1] <- 1
}
if(df_temp[Store_Code == j]$dist.km299 < 20) {
s[2] <- 1
}
if(df_temp[Store_Code == j]$dist.km300 < 20) {
s[3] <- 1
}
if(df_temp[Store_Code == j]$dist.km301 < 20) {
s[4] <- 1
}
}
v_temp <- data.table("Customer_ID" = c(i,i,i,i) , "Store_Code" =
c(60298,60299,60300,60301), "Prediction" = s)
l = list(vishal,v_temp)
vishal <- rbindlist(l)
}
)
dput(head(kk, 5))
structure(list(Customer_ID =
structure(c(1800000006365760, 1800000006365820,1800000006366060
,1800000006366060,1800000006366060), class = "integer64"), Store_Code =
c(60067, 60054, 60066,
60069, 60079), Purchase_2016 = c(2L, 1L, 1L, 1L, 2L), Purchase_2017 =
c(2L,
0L, 0L, 0L, 0L), TotalPurchases = c(4L, 1L, 1L, 1L, 2L), Return_2016 =
c(0L,
0L, 0L, 0L, 0L), Return_2017 = c(0L, 0L, 0L, 0L, 0L), Return_2010 = c(0L,
0L, 0L, 0L, 0L), Rp_Ratio_2016 = c(0, 0, 0, 0, 0), Rp_Ratio_2017 = c(0,
0, 0, 0, 0), Sales_Per_Day = c(1699.6, 2101.1, 1331.4, 1813.1,
1193.1), Store_Launch_Date = structure(c(1323820800, 1322006400,
1338163200, 1311984000, 1385164800), class = c("POSIXct", "POSIXt"
), tzone = "UTC"), Store_Size_Sq_Ft = c(8673.5, 12425.5, 15897.1,
6698.1, 3699.5), Customer_Count = c(89351, 118444, 79249, 114246,
54832), Total_Revenue = c(35350868.4, 43702303, 27693164.7, 37712369.7,
24816886.2), dist.km298 = c(140.24, 123.87, 10.2, 131.96, 128.52
), dist.km299 = c(163.37, 140.2, 79.32, 153.01, 145.03), dist.km300 =
c(4.09,
21.05, 126.55, 7.03, 17.41), dist.km301 = c(5.72, 19.04, 125.46,
5.02, 15.4), Nationality = c("INDIA", "UNITED ARAB EMIRATES",
"SRI LANKA", "SRI LANKA", "SRI LANKA"), Gender = c("M", "F",
"M", "M", "M"), Marital_Status = c("Married", "Married", "Married",
"Married", "Married"), Loyalty_Status = c("Gold", "Silver", "Silver",
"Silver", "Silver"), Points = c(814L, 212L, 186L, 186L, 186L),
Age = c(59L, 119L, 59L, 59L, 59L), LastVisit = c(2, 28, 3,
3, 3), Last_rdm_txn_dt1 = structure(c(17601, 16510, 17196,
17196, 17196), class = "Date"), Last_accr_txn_dt1 = structure(c(17801,
17029, 17774, 17774, 17774), class = "Date")), .Names = c("Customer_ID",
"Store_Code", "Purchase_2016", "Purchase_2017", "TotalPurchases",
"Return_2016", "Return_2017", "Return_2010", "Rp_Ratio_2016",
"Rp_Ratio_2017", "Sales_Per_Day", "Store_Launch_Date",
"Store_Size_Sq_Ft",
"Customer_Count", "Total_Revenue", "dist.km298", "dist.km299",
"dist.km300", "dist.km301", "Nationality", "Gender", "Marital_Status",
"Loyalty_Status", "Points", "Age", "LastVisit", "Last_rdm_txn_dt1",
"Last_accr_txn_dt1"), sorted = "Customer_ID", class = c("data.table",
"data.frame"), row.names = c(NA, -5L), .internal.selfref = <pointer:
0x0000000004810788>)

Answer 1

Please read the guide to posting good questions on StackOverflow, it will allow people to answer your questions easily. You question is very confusing but this might give you some ideas:
data %>%
group_by(Customer_ID, Store_Code) %>%
mutate(Predition298 = ifelse(dist.km298 > 20, 1, 0),
Predition299 = ifelse(dist.km299 > 20, 1, 0),
Predition300 = ifelse(dist.km300 > 20, 1, 0))