Section 3 Data Pre-processing

3.1 Read data into R

The National Chargepoint Register(NCR) is a database of publicly-available chargepoints for electric vehicles in the UK established in 2011(,2021). you can access data in this link

Now, let’s read original data in R

UK_NCR= read.csv(here::here("dataset","national-charge-point-registry.csv")) 
# This may take for a while, which depends on the speed of internet

# you can have a overview of this dataset
print("The number of rows is: ")
nrow(UK_NCR)
print("The number of columns is: ")
ncol(UK_NCR)
print("70 of all varriables are:")
head(names(UK_NCR),n = 70)

Tip: If you cannot successfully read this dataset, you can replace the above link with “https://raw.githubusercontent.com/Hereislittlemushroom/CASA0005_Final_Assessment/main/Dataset/national-charge-point-registry.csv”

3.2 Data Selection

Select the charge points of london area in this UK csv file。 You can utilise filter function from dplyr package to choose the charge point data in london boroughs

London_NCR = UK_NCR %>%
  dplyr::filter(  !is.na(county),
                  county == "London" | 
                  county == "Greater London " | 
                  county == "London Borough of Camden" |
                  county == "London Borough of Ealing" | 
                  county == "London Borough of Greenwich" | 
                  county == "London Borough of Hackney" | 
                  county == "London Borough of Hammersmith and Fulham" | 
                  county == "London Borough of Hounslow" | 
                  county == "London Borough of Islington" | 
                  county == "London Borough of Lambeth" | 
                  county == "London Borough of Richmond upon Thames" | 
                  county == "London Borough of Southwark" |
                  county == "London Borough Of Southwark" |
                  county == "London Borough of Waltham Forest" | 
                  county == "London Borough of Wandsworth")

Check if all values in county are attributed to “London”

isLondon = London_NCR$county %>%
  unique()
isLondon

In the next step, you can select the valuable attributes e.g. latitude,longitude.

# Tip: the index of data frame starts from 1
# Select the variables by their index
London_NCR = London_NCR %>%
  select(1,4,5,13,14,15,32,35,36,38,54)

# Check the variables we have chosen and the number of rows & cols
London_NCR %>%
  names()
London_NCR %>%
  nrow()
London_NCR %>%
  ncol()

3.3 Data Cleaning

Map and visualisation play important roles in spatial analysis. To make a heat map for further research, you need to merge geographic information for each row in charge point dataset in the first place.

To begin with, import “PostcodesioR.” This R package offer methods to match

# install.packages("PostcodesioR")
library(PostcodesioR)

Before applying “for-loop” method to fill values in GSS_CODE by identifying postcode, you can add a new columns called GSS_CODE in London_NCR dataset.

# Attentions: you can skip this chunk because the for-loop process can take for a quite long time (about 5 min).
# It is not necessary to stick on it, just skip!

London_NCR_GSS_Added = London_NCR %>%
  rowwise() %>%
  mutate(GSS_CODE = postcode) %>%
  # Tip: it is essential to transform numerical data into one in character
  mutate(GSS_CODE = as.character(GSS_CODE))

# Pay attention to the for loop in dataframe, it starts from 1

i = 1
for (val in London_NCR_GSS_Added$postcode) {
  try({ temp1 = PostcodesioR::postcode_lookup(val)
        if(!is.null(temp)){
          temp2 = temp1$admin_district_code[1]
          London_NCR_GSS_Added$GSS_CODE[i] = temp2
        }else{
          London_NCR_GSS_Added$GSS_CODE[i] = ""
        }
        i = i+1 }
      ,silent = TRUE)
}

# remove the rows whose value of `GSS_CODE` is empty
# There are limitations in this process because the rows missing `GSS_CODE` cannot be included in the dataset, which can slightly affect the research results 

London_NCR_GSS_Added$GSS_CODE[London_NCR_GSS_Added$GSS_CODE==""] = NA
London_NCR_GSS_Added = London_NCR_GSS_Added %>%
  filter(!is.na(GSS_CODE))

Finally, it is of importance to export our prepossessed data into csv file! Now we get the London_NCR_GSS_Added.csv in our “/Dataset” path.

# export London_NCR_GSS_Added data frame into .CSV format
library(here)
write.csv(London_NCR_GSS_Added, here::here("Dataset","London_NCR_GSS_Added.csv"), row.names = FALSE, col.names = TRUE)
# `col.names = TRUE` is important to be writen down

Also, you can access this prepocessed dataset in github link: https://raw.githubusercontent.com/Hereislittlemushroom/CASA0005_Final_Assessment/main/Dataset/London_NCR_GSS_Added.csv