In phillipf/GNAFAPI:
library(dplyr)
library(data.table)
# library(devtools)
# install_github("phillipf/AddressR")
library(AddressR)
library(RODBC)
library(sp)
library(stringdist)
CoolingTowers <- read.csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/050117_DHHSCTSAreadetails.csv',
stringsAsFactors=FALSE)
CoolingTower_Sites <- data.frame(SITE = CoolingTowers$X.1[grepl("SITE", CoolingTowers$X.1)],
Geocode = CoolingTowers$X.1[grepl("Geocoded", CoolingTowers$X.1)]) %>%
mutate(x = sapply(strsplit(sub("Geocoded: ", "", Geocode), "/"), "[", 1),
coord = sapply(strsplit(sub("Geocoded: ", "", Geocode), "/"), "[", 2)) %>%
transmute(SITE,
LATITUDE = unname(x),
LONGITUDE = sapply(
sapply(coord, strsplit, '[(]'),
"[", 1),
ACCURACY = sapply(sapply(coord, strsplit, '[(]'),
"[", 2)) %>%
mutate(ADDRESS_STRING = gsub("(SITE\\s*[0-9]+)\\s*-\\s*(.*)","\\2",SITE),
SITE = as.numeric(gsub("SITE\\s*([0-9]+)\\s*-\\s*(.*)","\\1",SITE)),
ACCURACY = gsub("\\)$", "", ACCURACY))
write.csv(CoolingTower_Sites,
'N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv',
row.names = F)
paste3 <- function(...,sep=" ") {
L <- list(...)
L <- lapply(L,function(x) {x[is.na(x)] <- ""; x})
ret <-gsub(paste0("(^",sep,"|",sep,"$)"),"",
gsub(paste0(sep,sep),sep,
do.call(paste,c(L,list(sep=sep)))))
is.na(ret) <- ret==""
ret
}
EMIS <- read.csv('N:/ABR/TradeWasteCustomers/15062017_ActiveInactive.csv',
stringsAsFactors=FALSE,
skip=6) %>%
select(AG_CONSUMERNUMBER, textbox11, textbox65, textbox43, textbox46, textbox49, textbox53, textbox56) %>% mutate(MASTERID = as.numeric(substr(AG_CONSUMERNUMBER, 3, 9)))
EMIS$ADDRESS_STRING <- paste3(paste3(EMIS$textbox11, EMIS$textbox65, EMIS$textbox43), paste3(EMIS$textbox46, EMIS$textbox49), paste3(EMIS$textbox53, "VIC", EMIS$textbox56), sep = ", ")
write.csv(EMIS, "N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_unlabelled_EMIS.csv",
row.names = F)
Gentrack <- read.csv('file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/Metersequencesdetail_June17.csv',
stringsAsFactors=FALSE,
skip=3)
unlabelled_gentrack <- Gentrack %>%
select(INSTALL, INSTALL_ADDRESS) %>%
distinct() %>%
mutate(MASTERID = as.integer(substr(as.character(INSTALL), 3, 9))) %>%
filter(MASTERID %in% EMIS$MASTERID) %>%
mutate(MASTERID = as.character(MASTERID))
write.csv(unlabelled_gentrack, "N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_unlabelled_Gentrack.csv",
row.names = F)
```{python DHHS parserator}
import re
import GNAF
import itertools
import operator
import pandas as pd
unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv", header=None)
site = unlabelled[0]
unlabelled = unlabelled[4].apply(lambda x: re.sub(r', Australia$', "", x))
def accumulate(l):
it = itertools.groupby(l, operator.itemgetter(1))
for key, subiter in it:
yield key, ' '.join(item[0] for item in subiter)
labelled = unlabelled.apply(lambda x: GNAF.parse(x))
labelled = labelled.apply(lambda x: list(accumulate(x)))
cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX',
'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME',
'POSTCODE', 'STATE_ABBREVIATION']
df = pd.DataFrame(columns=cols)
df['SITE'] = site
df['ADDRESS_STRING'] = unlabelled
for i, row in enumerate(labelled):
for item in row:
df[item[0]][i] = item[1]
df2 = df.ix[1:]
df2.to_csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_labelled_DHHS.csv', index=False)
```{python EMIS parserator}
import re
import GNAF
import itertools
import operator
import pandas as pd
unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_unlabelled_EMIS.csv", header=None)
site = unlabelled[0]
unlabelled = unlabelled[8].apply(lambda x: re.sub(r', Australia$', "", x))
def accumulate(l):
it = itertools.groupby(l, operator.itemgetter(1))
for key, subiter in it:
yield key, ' '.join(item[0] for item in subiter)
labelled = unlabelled.apply(lambda x: GNAF.parse(x))
labelled = labelled.apply(lambda x: list(accumulate(x)))
cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX',
'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME',
'POSTCODE', 'STATE_ABBREVIATION']
df = pd.DataFrame(columns=cols)
df['SITE'] = site
df['ADDRESS_STRING'] = unlabelled
for i, row in enumerate(labelled):
for item in row:
df[item[0]][i] = item[1]
df2 = df.ix[1:]
df2.to_csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_labelled_EMIS.csv', index=False)
```{python Gentrack parserator}
import re
import GNAF
import itertools
import operator
import pandas as pd
unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_unlabelled_Gentrack.csv", header=None)
site = unlabelled[2]
unlabelled = unlabelled[1]
def accumulate(l):
it = itertools.groupby(l, operator.itemgetter(1))
for key, subiter in it:
yield key, ' '.join(item[0] for item in subiter)
labelled = unlabelled.apply(lambda x: GNAF.parse(x))
labelled = labelled.apply(lambda x: list(accumulate(x)))
cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX',
'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME',
'POSTCODE', 'STATE_ABBREVIATION']
df = pd.DataFrame(columns=cols)
df['SITE'] = site
df['ADDRESS_STRING'] = unlabelled
for i, row in enumerate(labelled):
for item in row:
df[item[0]][i] = item[1]
df2 = df.ix[1:]
df2.to_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_labelled_Gentrack.csv", index=False)
```r
labelled_EMIS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/160517_labelled_EMIS.csv", stringsAsFactors = F)
manual_training <- labelled_EMIS %>%
filter(BUILDING_NAME != "") %>%
sample_frac(0.2)
manual_training2 <- labelled_EMIS %>%
filter(BUILDING_NAME == "") %>%
sample_frac(0.005)
manual_training3 <- rbind(manual_training, manual_training2) %>%
select(ADDRESS_STRING)
write.table(manual_training3, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS.csv",
row.names = F, col.names = F)
manual_training <- labelled_EMIS %>%
filter(BUILDING_NAME == "LOT" |
FLAT_TYPE == "PREV" |
grepl("SH", BUILDING_NAME) |
grepl("GRN", BUILDING_NAME) |
grepl("GRD", BUILDING_NAME) |
grepl("LOWERGROUND", BUILDING_NAME) |
grepl("GROUND", FLAT_TYPE) |
FLAT_TYPE == "GR" |
FLAT_TYPE == "G") %>%
select(ADDRESS_STRING)
write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS2.csv",
row.names = F, col.names = F)
manual_training <- labelled_EMIS %>%
filter(BUILDING_NAME == "PRINCES HIGHWAY," |
BUILDING_NAME == "WELLINGTON PARADE," |
BUILDING_NAME == "OLYMPIC BOULEVARD," |
BUILDING_NAME == "PARK CRESCENT," |
BUILDING_NAME == "MELBOURNE PLACE,") %>%
select(ADDRESS_STRING)
write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS3.csv",
row.names = F, col.names = F)
manual_training <- labelled_EMIS %>%
filter(grepl("^[A-Z]{1}\\s.*", STREET_NAME)) %>%
select(ADDRESS_STRING)
write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/090617_unlabelled_EMIS4.csv",
row.names = F, col.names = F)
manual_training <- labelled_EMIS %>%
filter(grepl("MELBOURNE UNIVERSITY", ADDRESS_STRING)) %>%
select(ADDRESS_STRING)
write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/200617_unlabelled_EMIS5.csv",
row.names = F, col.names = F)
ABR <- odbcDriverConnect(connection="Driver={SQL Server Native Client 11.0};
server= WVdb1devsql;
database=ABR;
uid=CWW\\farrelp1;
Trusted_Connection=Yes;")
GNAF_address <- sqlQuery(ABR,
"SELECT DISTINCT
[ADDRESS_DETAIL_PID]
,[BUILDING_NAME]
,CASE
WHEN [NUMBER_FIRST] % 2 = 0
THEN REPLACE(SUBSTRING([STREET_NAME], 1, 1) + 'EVEN', ' ', '')
ELSE REPLACE(SUBSTRING([STREET_NAME], 1, 1) + 'ODD', ' ', '')
END AS [KEY]
,[LOT_NUMBER]
,[NUMBER_FIRST]
,[NUMBER_LAST]
,[STREET_NAME]
,[STREET_TYPE_CODE]
,CASE
WHEN [STREET_TYPE_CODE] IS NULL
THEN [STREET_NAME]
ELSE [STREET_NAME] + ' ' + [STREET_TYPE_CODE]
END AS [STREET]
,[LOCALITY_NAME]
,[POSTCODE]
,[LONGITUDE]
,[LATITUDE]
,CASE
WHEN [NUMBER_LAST] IS NULL
THEN 1
ELSE 2
END AS ADDRESS_TYPE
,CONFIDENCE
,[PRIMARY_SECONDARY]
FROM [ABR].[dbo].[ADDRESS_VIEW]
WHERE [PRIMARY_SECONDARY] = 'P' OR [PRIMARY_SECONDARY] IS NULL;",
stringsAsFactors=FALSE)
text_clean <- function(x, field.type = "text") {
if(x == "" | is.na(x)) {
return(NA)
}
else if (field.type == "text"){
x <- gsub(",$","",x)
return(toupper(x))
}
else if (field.type == "num"){
return(as.numeric(gsub(",","",x)))
}
}
text_clean <- Vectorize(text_clean)
labelled_DHHS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_labelled_DHHS.csv", stringsAsFactors = F)
DHHS_sites <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv", stringsAsFactors = F)
indx <- !is.na(labelled_DHHS$NUMBER_FIRST_SUFFIX) & is.numeric(labelled_DHHS$NUMBER_FIRST_SUFFIX)
labelled_DHHS[indx,]$NUMBER_FIRST <- labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX
labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX <- NA
indx <- grepl("[A-Z]$", labelled_DHHS$NUMBER_FIRST)
numFirst <- gsub("([0-9]+)[A-Z]", "\\1", labelled_DHHS[indx,]$NUMBER_FIRST)
suffix <- gsub("([0-9]+)([A-Z])", "\\2", labelled_DHHS[indx,]$NUMBER_FIRST)
labelled_DHHS[indx,]$NUMBER_FIRST <- numFirst
labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX <- suffix
labelled_DHHS <- labelled_DHHS %>%
mutate(BUILDING_NAME = text_clean(BUILDING_NAME),
LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")),
LEVEL_TYPE = text_clean(LEVEL_TYPE),
LEVEL_NUMBER = text_clean(LEVEL_NUMBER, field.type = "num"),
FLAT_TYPE = text_clean(FLAT_TYPE),
FLAT_NUMBER_PREFIX = text_clean(FLAT_NUMBER_PREFIX),
FLAT_NUMBER = text_clean(FLAT_NUMBER, field.type = "num"),
NUMBER_FIRST = text_clean(NUMBER_FIRST, field.type = "num"),
NUMBER_LAST = text_clean(NUMBER_LAST, field.type = "num"),
STREET_NAME = text_clean(STREET_NAME),
STREET_TYPE_CODE = text_clean(STREET_TYPE_CODE),
LOCALITY_NAME = text_clean(LOCALITY_NAME),
POSTCODE = text_clean(POSTCODE, field.type = "num"),
STATE_ABBREVIATION = text_clean(STATE_ABBREVIATION),
KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3),
substr(gsub("'","", STREET_NAME), 1,1),
ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))),
ID = row_number(),
ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2)) %>%
left_join(select(DHHS_sites, -ADDRESS_STRING), by = "SITE")
GNAF_address <- GNAF_address %>%
mutate(PID = as.numeric(gsub("GAVIC", "", ADDRESS_DETAIL_PID)))
Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.4.cpp")
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.1.cpp")
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev7.cpp")
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev9.cpp")
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev10.cpp")
Clean2 <- function(nonblanks, ID) {
# GNAF_address <- data.table::fread("C:/Users/farrelp1/Documents/GNAFAPI/data/May17_GNAF.csv", stringsAsFactors = F)
#
# colnames(GNAF_address) <- colnames(data.table::fread("C:/Users/farrelp1/Documents/AddressR/data/May17_GNAF_address.csv", stringsAsFactors = F))[1:15]
options <- lapply(nonblanks$KEY, function(x) which(x == GNAF_address$KEY))
names(options) <- ID
totalscore <- lapply(seq_along(options), function(i) parallelAddressTotalScore6(nonblanks[i,], GNAF_address[options[[i]],]))
names(totalscore) <- ID
options2 <- lapply(seq_along(options), function(i) cbind(option = suppressWarnings(options[[i]][which(totalscore[[i]] == min(totalscore[[i]]))]),
score = suppressWarnings(totalscore[[i]][which(totalscore[[i]] == min(totalscore[[i]]))])))
names(options2) <- names(options)
options3 <- options2[unname(unlist(lapply(options2, function(x) nrow(x) > 0)))]
finalresult <- plyr::ldply(names(options3), function(i) cbind(id = i,options3[[i]], GNAF_address[options2[[i]][,1],]))
# finalresult <- plyr::ldply(names(options3), function(i) cbind(id = i,options3[[i]], GNAF_address[options2[[i]][,1],])) %>%
# mutate(id = as.numeric(as.character(id))) %>%
# group_by(id) %>%
# filter(CONFIDENCE == max(CONFIDENCE))
#finalresult <- plyr::ldply(options4, function(x) x %>% filter(CONFIDENCE == max(CONFIDENCE))) #%>%
#mutate(ID = as.numeric(.id))
#dup <- finalresult[duplicated(finalresult$.id),]
#CWW_TradeWaste_ABR <- select(finalresult, AG_CONSUMERNUMBER, ADDRESS_DETAIL_PID) %>% distinct()
return(finalresult)
}
Clean3 <- function(nonblanks, ID) {
GNAF_address2 <- data.table::as.data.table(GNAF_address)
setkey(GNAF_address2, "KEY")
nonblanks2 <- data.table::as.data.table(nonblanks)
setkey(nonblanks2, "KEY")
totalscore2 <- lapply(1:nrow(nonblanks2), function(i) {
if(!is.na(nonblanks2[[i, "NUMBER_FIRST"]])) {
options <- GNAF_address2[GNAF_address2[["KEY"]] == nonblanks2[[i, "KEY"]] & GNAF_address2[["NUMBER_FIRST"]] < nonblanks2[[i, "NUMBER_FIRST"]] + 50 & GNAF_address2[["NUMBER_FIRST"]] > nonblanks2[[i, "NUMBER_FIRST"]] - 50]
}
else if(!is.na(nonblanks2[[i, "LOT_NUMBER"]])) {
options <- GNAF_address2 %>% filter(!is.na(LOT_NUMBER))
}
else {
# options <- GNAF_address2[gsub("([A-Z])(.*)", "\\1", GNAF_address2[["KEY"]]) == gsub("([A-Z])(.*)", "\\1", nonblanks2[[i, "KEY"]])]
options <- GNAF_address2 %>% filter(!is.na(BUILDING_NAME))
}
result <- parallelAddressTotalScore6(nonblanks2[i,], options)
colnames(result) = options[["ADDRESS_DETAIL_PID"]]
return(result)
})
names(totalscore2) <- nonblanks2[["SITE"]]
setkey(GNAF_address2, "ADDRESS_DETAIL_PID")
finalresult <- plyr::ldply(names(totalscore2), function(i) {
if(length(totalscore2[[i]]) > 0) {
ADDRESS_DETAIL_PID <- names(totalscore2[[i]][,which.min(totalscore2[[i]])])
score <- as.numeric(unname(totalscore2[[i]][,which.min(totalscore2[[i]])]))
result <- as.data.table(cbind(id = as.numeric(i), ADDRESS_DETAIL_PID = ADDRESS_DETAIL_PID, score = score))
setkey(result, "ADDRESS_DETAIL_PID")
return(GNAF_address2[result])
}
else{}
})
finalresult$score <- as.numeric(finalresult$score)
return(finalresult)
}
#AddressR::AddressClean2
#DHHS_GNAF <- AddressR::AddressClean2(nonblanks=labelled_DHHS, ID=labelled_DHHS$ID)
DHHS_GNAF <- Clean3(nonblanks=labelled_DHHS, ID=labelled_DHHS$SITE)
# for (i in seq_along(options)) {
#
# r <- parallelAddressTotalScore3(nonblanks[i, ], GNAF_address[options[[i]], ])
#
# }
# test <- parallelAddressTotalScore9(labelled_DHHS[1,], GNAF_address) works but very slow
#test <- parallelAddressTotalScore10(labelled_DHHS[1:2,], GNAF_address)
text_clean <- function(x, field.type = "text") {
if(x == "" | is.na(x)) {
return(NA)
}
else if (field.type == "text"){
x <- gsub(",$","",x)
return(toupper(x))
}
else if (field.type == "num"){
return(as.numeric(gsub(",","",x)))
}
}
text_clean <- Vectorize(text_clean)
EMIS_sites <- read.csv('N:/ABR/TradeWasteCustomers/15062017_ActiveInactive.csv',
stringsAsFactors=FALSE,
skip=6) %>%
select(AG_CONSUMERNUMBER, AG_COMPANYNAME, AG_TRADINGNAME, textbox11, textbox65, textbox43, textbox46, textbox49, textbox53, textbox56, PLD_PROPERTYXYCOORDINATE) %>%
mutate(AG_COMPANYNAME = ifelse(AG_TRADINGNAME == "", AG_COMPANYNAME, AG_TRADINGNAME),
SITE = AG_CONSUMERNUMBER,
LAT=as.integer(gsub("([0-9]{6})([0-9]{7})","\\1",PLD_PROPERTYXYCOORDINATE)),
LON=as.integer(gsub("([0-9]{6})([0-9]{7})","\\2",PLD_PROPERTYXYCOORDINATE)))
CWWBoundary <- rgdal::readOGR("N:/Asset Information/MUNSYS MapInfo Data/Production/Data/CWW Boundary_2014_region.shp",
layer="CWW Boundary_2014_region")
EMIS_sp <- EMIS_sites %>%
filter(!is.na(LAT) & LAT != 0 & LAT != 999999)
EMIS_sp <- sp::SpatialPointsDataFrame(coords=data.frame(X=EMIS_sp$LAT, Y=EMIS_sp$LON),
data=data.frame(SITE=EMIS_sp$SITE),
proj4string=sp::CRS(proj4string(CWWBoundary)))
EMIS_sp <- sp::spTransform(EMIS_sp, sp::CRS("+init=epsg:4326 +proj=longlat"))
EMIS_sites <- EMIS_sites %>%
left_join(data.frame(SITE=EMIS_sp@data,
LATITUDE = EMIS_sp@coords[,2],
LONGITUDE = EMIS_sp@coords[,1]))
labelled_EMIS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_labelled_EMIS.csv", stringsAsFactors = F) %>%
#left_join(select(EMIS_sites, SITE, AG_COMPANYNAME)) %>%
mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>%
mutate(#BUILDING_NAME = ifelse(BUILDING_NAME == "", AG_COMPANYNAME, BUILDING_NAME),
MASTERID = as.integer(substr(MASTERID, 3, 9)))
indx <- !is.na(labelled_EMIS$NUMBER_FIRST_SUFFIX) & is.numeric(labelled_EMIS$NUMBER_FIRST_SUFFIX)
if(any(indx == TRUE)) {
labelled_EMIS[indx,]$NUMBER_FIRST <- labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX
labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- NA
}
indx <- grepl("[A-Z]$", labelled_EMIS$NUMBER_FIRST)
if(any(indx == TRUE)) {
numFirst <- gsub("([0-9]+)[A-Z]", "\\1", labelled_EMIS[indx,]$NUMBER_FIRST)
suffix <- gsub("([0-9]+)([A-Z])", "\\2", labelled_EMIS[indx,]$NUMBER_FIRST)
labelled_EMIS[indx,]$NUMBER_FIRST <- numFirst
labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- suffix
}
indx <- grepl("^[A-Z]{1}\\s.*", labelled_EMIS$STREET_NAME)
if(any(indx == TRUE)) {
suffix <- gsub("^([A-Z]{1})\\s(.*)", "\\1", labelled_EMIS[indx,]$STREET_NAME)
streetname <- gsub("^([A-Z]{1})\\s(.*)", "\\2", labelled_EMIS[indx,]$STREET_NAME)
labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- suffix
labelled_EMIS[indx,]$STREET_NAME <- streetname
}
indx <- grepl("^MOUNT\\s.*", labelled_EMIS$STREET_NAME)
labelled_EMIS$STREET_NAME[indx] <- gsub("^(MOUNT)", "MT", labelled_EMIS[indx,]$STREET_NAME)
indx <- grepl("^17C\\s.*", labelled_EMIS$STREET_NAME)
labelled_EMIS$STREET_NAME[indx] <- gsub("^(17C\\s)(.*)", "\\2", labelled_EMIS[indx,]$STREET_NAME)
labelled_EMIS <- labelled_EMIS %>%
mutate(BUILDING_NAME = unname(text_clean(BUILDING_NAME)),
LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")),
LEVEL_TYPE = unname(text_clean(LEVEL_TYPE)),
LEVEL_NUMBER = unname(text_clean(LEVEL_NUMBER, field.type = "num")),
FLAT_TYPE = unname(text_clean(FLAT_TYPE)),
FLAT_NUMBER_PREFIX = unname(text_clean(FLAT_NUMBER_PREFIX)),
FLAT_NUMBER = unname(text_clean(FLAT_NUMBER, field.type = "num")),
NUMBER_FIRST = unname(text_clean(NUMBER_FIRST, field.type = "num")),
NUMBER_LAST = unname(text_clean(NUMBER_LAST, field.type = "num")),
STREET_NAME = unname(text_clean(STREET_NAME)),
STREET_TYPE_CODE = unname(text_clean(STREET_TYPE_CODE)),
LOCALITY_NAME = unname(text_clean(LOCALITY_NAME)),
POSTCODE = unname(text_clean(POSTCODE, field.type = "num")),
STATE_ABBREVIATION = unname(text_clean(STATE_ABBREVIATION)),
KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3),
substr(gsub("'","", STREET_NAME), 1,1),
ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))),
#ID = row_number(),
ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2)) #%>%
#left_join(labelled_Gentrack, by = "MASTERID")
labelled_Gentrack <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_labelled_Gentrack.csv", stringsAsFactors = F) %>%
mutate(MASTERID = SITE)
labelled_Gentrack <- labelled_Gentrack %>%
mutate(BUILDING_NAME = unname(text_clean(BUILDING_NAME)),
LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")),
LEVEL_TYPE = unname(text_clean(LEVEL_TYPE)),
LEVEL_NUMBER = unname(text_clean(LEVEL_NUMBER, field.type = "num")),
FLAT_TYPE = unname(text_clean(FLAT_TYPE)),
FLAT_NUMBER_PREFIX = unname(text_clean(FLAT_NUMBER_PREFIX)),
FLAT_NUMBER = unname(text_clean(FLAT_NUMBER, field.type = "num")),
NUMBER_FIRST = unname(text_clean(NUMBER_FIRST, field.type = "num")),
NUMBER_LAST = unname(text_clean(NUMBER_LAST, field.type = "num")),
STREET_NAME = unname(text_clean(STREET_NAME)),
STREET_TYPE_CODE = unname(text_clean(STREET_TYPE_CODE)),
LOCALITY_NAME = unname(text_clean(LOCALITY_NAME)),
POSTCODE = unname(text_clean(POSTCODE, field.type = "num")),
STATE_ABBREVIATION = unname(text_clean(STATE_ABBREVIATION)),
KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3),
substr(gsub("'","", STREET_NAME), 1,1),
ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))),
#ID = row_number(),
ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2))
# labelled_Gentrack <- labelled_Gentrack %>%
# left_join(select(mutate(Gentrack,
# MASTERID = as.integer(substr(as.character(INSTALL), 3, 9))),
# MASTERID,
# AG_COMPANYNAME)) %>%
# #mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>%
# mutate(BUILDING_NAME = ifelse(is.na(BUILDING_NAME), AG_COMPANYNAME, BUILDING_NAME))
labelled_EMIS2 <- labelled_EMIS
lapply(1:nrow(labelled_EMIS), function(i) {
NUMBER_FIRST = labelled_EMIS[i,]$NUMBER_FIRST
LOT_NUMBER = labelled_EMIS[i,]$LOT_NUMBER
MASTERID = labelled_EMIS[i,]$MASTERID
REPLACEMENT = labelled_Gentrack[labelled_Gentrack$MASTERID == MASTERID,] #%>%
#filter(!is.na(NUMBER_FIRST)) %>%
#distinct()
if(nrow(REPLACEMENT) > 0) {
if(is.na(NUMBER_FIRST) & is.na(LOT_NUMBER)) {
labelled_EMIS2 <<- rbind(labelled_EMIS2, REPLACEMENT)
labelled_EMIS2 <<- labelled_EMIS2[-i,]
}
}
})
labelled_EMIS2 <- labelled_EMIS2 %>%
left_join(select(EMIS_sites, SITE, AG_COMPANYNAME, LATITUDE, LONGITUDE)) %>%
#mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>%
mutate(BUILDING_NAME = ifelse(is.na(BUILDING_NAME), AG_COMPANYNAME, BUILDING_NAME)) %>%
distinct()
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.cpp")
Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.4.cpp")
# labelled_EMIS2$LATITUDE = rep(as.numeric(NA), length=nrow(labelled_EMIS2))
#
# labelled_EMIS2$LONGITUDE = rep(as.numeric(NA), length=nrow(labelled_EMIS2))
# sample <- labelled_EMIS %>% mutate(SITE=as.numeric(SITE)) %>% filter(!is.na(BUILDING_NAME) & is.na(NUMBER_FIRST)) %>% sample_n(10)
EMIS_GNAF <- Clean3(nonblanks=labelled_EMIS2, ID=labelled_EMIS2 %>% distinct() %>% .[["SITE"]])
DHHS_GNAF2 <- DHHS_GNAF %>%
select(ADDRESS_DETAIL_PID, id, NUMBER_FIRST:LATITUDE) %>%
mutate(id = as.numeric(id)) %>%
#left_join(CoolingTower_Sites, by=c("id" = "SITE")) %>%
left_join(select(labelled_DHHS, SITE, BUILDING_NAME), by=c("id" = "SITE")) %>%
as.data.table()
setkey(DHHS_GNAF2, "ADDRESS_DETAIL_PID")
#EMIS_GNAF2 <- as.data.table(EMIS_GNAF)
EMIS_GNAF2 <- EMIS_GNAF %>%
select(ADDRESS_DETAIL_PID, id) %>%
mutate(id = as.numeric(id)) %>%
#left_join(CoolingTower_Sites, by=c("id" = "SITE")) %>%
#left_join(select(labelled_EMIS, SITE, BUILDING_NAME), by=c("id" = "SITE")) %>%
as.data.table()
setkey(EMIS_GNAF2, "ADDRESS_DETAIL_PID")
overlay <- EMIS_GNAF2[DHHS_GNAF2]
# options <- EMIS_GNAF[EMIS_GNAF$STREET == overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$STREET,]
#
# options$EVEN <- ifelse(options$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD")
#
# overlay$EVEN <- ifelse(overlay$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD")
#
# match <- options[overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$NUMBER_FIRST <= options$NUMBER_LAST & overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$NUMBER_FIRST >= options$NUMBER_FIRST & overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$EVEN == options$EVEN,]$ADDRESS_DETAIL_PID
#
# match <- unique(match[!is.na(match)])
#
# names(match) <- rep(overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$i.id, length(match))
#
# match <- plyr::ldply(match, function(x) c(ADDRESS_DETAIL_PID = unname(x), SITE = names(x))) %>%
# left_join(EMIS_GNAF) %>%
# .[["id"]]
overlay$EVEN <- ifelse(overlay$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD")
range_check <- function(ID, STREET, NUMBER_FIRST, EVEN, EMIS_GNAF) {
if(is.na(ID)) {
options <- EMIS_GNAF[EMIS_GNAF$STREET == STREET,]
if(nrow(options) > 0) {
options$EVEN <- ifelse(options$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD")
match <- options[NUMBER_FIRST <= options$NUMBER_LAST & NUMBER_FIRST >= options$NUMBER_FIRST & EVEN == options$EVEN,]$ADDRESS_DETAIL_PID
match2 <- unique(match[!is.na(match)])
#names(match2) <- rep("ADDRESS_DETAIL_PID", length(match2))
if(length(match2) > 0) {
match3 <- plyr::ldply(match2) %>%
left_join(EMIS_GNAF, by = c("V1" = "ADDRESS_DETAIL_PID")) %>%
.[["id"]]
return(match3)
}
}
}
}
results <- sapply(1:nrow(overlay), function(x) range_check(overlay$id[x], overlay$STREET[x], overlay$NUMBER_FIRST[x], overlay$EVEN[x], EMIS_GNAF))
names(results) <- overlay$i.id
# df <- data.frame(matrix(unlist(results), nrow=length(unlist(results[unlist(lapply(results, function(x) length(x) > 0))])), byrow=T))
df <- plyr::ldply(lapply(seq_along(results), function(i) data.frame(CONSUMER = results[[i]],
i.id = rep(names(results[i]), length(results[[i]])), stringsAsFactors = F))) %>%
mutate(i.id = as.numeric(i.id))
#do( data.frame(column = rep(results, each = length(results)), stringsAsFactors = FALSE) )
overlay <- overlay %>%
left_join(df) %>%
mutate(id = ifelse(is.na(id), CONSUMER, id)) %>%
select(ADDRESS_DETAIL_PID:BUILDING_NAME) %>%
filter(is.na(id)) %>%
distinct()
# VICMAP_LandUse <- readOGR("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/2. Zones/ll_gda94/plan_overlay.shp",
# layer="plan_overlay")
gap <- sp::SpatialPointsDataFrame(coords = data.frame(X=overlay$LONGITUDE, Y=overlay$LATITUDE),
data=data.frame(id = overlay$i.id),
proj4string=sp::CRS("+init=epsg:4326 +proj=longlat")) %>%
sp::spTransform(sp::CRS(sp::proj4string(CWWBoundary)))
#VICMAP_LandUse2 <- spTransform(VICMAP_LandUse, CRS(proj4string(CWWBoundary)))
#proj4string(CWWBoundary)
gap_cww <- gap[which(rgeos::gContains(CWWBoundary, gap, byid = TRUE)), ]
gap_cww <- gap_cww@data %>%
left_join(select(overlay, 1, i.id:BUILDING_NAME), by = c("id"= "i.id"))
# STREET_SCORE <- lapply(gap_cww, function(x) {
# if(names(x) == "STREET") {
# STREET_SCORE <- stringdist(x$STREET, EMIS_GNAF$STREET)
# }
# #SUBURB_SCORE <- stringdist(x$LOCALITY_NAME, EMIS_GNAF$LOCALITY_NAME)
# })
#
#
# idx <- lapply(gap_cww, is.character)
#
# gap_cww[,unname(unlist(idx))][,-1]
#
# idx <- lapply(EMIS_GNAF, is.character)
#
# EMIS_GNAF[,unname(unlist(idx))][,!c("KEY", "PRIMARY_SECONDARY", "id")]
gap <- data.table(gap_cww)
add <- function(x) Reduce("+", x)
subtract <- function(x) Reduce("-", x)
EMIS2 <- data.table(EMIS_GNAF)
cnames <- c("BUILDING_NAME", "STREET", "LOCALITY_NAME")
score <- list()
for(cname in cnames) {
score[[cname]] <- stringdistmatrix(gap[[cname]], EMIS2[[cname]])
}
result <- matrix(nrow = nrow(gap), ncol = nrow(EMIS2))
numdiff <- function(x, y) {abs(x - y)}
#test <- Map(numdiff, gap[["NUMBER_FIRST"]], EMIS2[["NUMBER_FIRST"]])
#x <- foreach(i=gap[["NUMBER_FIRST"]]) %do% numdiff(i, EMIS2[["NUMBER_FIRST"]])
#x <- foreach(i=EMIS2[["NUMBER_FIRST"]], .combine='rbind') %do% numdiff(i, gap[["NUMBER_FIRST"]])
cnames <- c("NUMBER_FIRST", "NUMBER_LAST")
for(cname in cnames) {
score[[cname]] <- foreach(i=EMIS2[[cname]], .combine='cbind') %do% numdiff(i, gap[[cname]])
dimnames(score[[cname]]) <- NULL
}
#apply(gap, c(1,2), function(x) )
# for(cname in cnames) {
#
# for(i in 1:nrow(gap)) {
#
# for(j in 1:nrow(EMIS2))
#
# #if(!is.na(gap[[cname]][i]))
#
# result[i,j] <- ifelse(is.na(abs(gap[[cname]][i] - EMIS2[[cname]][j])), 0, abs(gap[[cname]][i] - EMIS2[[cname]][j]))
#
# }
#
# score[[cname]] <- result
#
# }
#subtract(list(gap[["NUMBER_FIRST"]], EMIS2[["NUMBER_FIRST"]]))
score2 <- add(score)
score3 <- apply(score2, 1, which.min)
closest_match <- data.table(gap[["id"]], EMIS2[score3,])
write.csv(EMIS_GNAF, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_EMIS_GNAF.csv",
row.names = F)
write.csv(DHHS_GNAF, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_DHHS_GNAF.csv",
row.names = F)
write.csv(gap_cww, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_DHHS_gap.csv",
row.names = F)
phillipf/GNAFAPI documentation built on May 25, 2019, 5:05 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
library(dplyr) library(data.table) # library(devtools) # install_github("phillipf/AddressR") library(AddressR) library(RODBC) library(sp) library(stringdist)
CoolingTowers <- read.csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/050117_DHHSCTSAreadetails.csv', stringsAsFactors=FALSE) CoolingTower_Sites <- data.frame(SITE = CoolingTowers$X.1[grepl("SITE", CoolingTowers$X.1)], Geocode = CoolingTowers$X.1[grepl("Geocoded", CoolingTowers$X.1)]) %>% mutate(x = sapply(strsplit(sub("Geocoded: ", "", Geocode), "/"), "[", 1), coord = sapply(strsplit(sub("Geocoded: ", "", Geocode), "/"), "[", 2)) %>% transmute(SITE, LATITUDE = unname(x), LONGITUDE = sapply( sapply(coord, strsplit, '[(]'), "[", 1), ACCURACY = sapply(sapply(coord, strsplit, '[(]'), "[", 2)) %>% mutate(ADDRESS_STRING = gsub("(SITE\\s*[0-9]+)\\s*-\\s*(.*)","\\2",SITE), SITE = as.numeric(gsub("SITE\\s*([0-9]+)\\s*-\\s*(.*)","\\1",SITE)), ACCURACY = gsub("\\)$", "", ACCURACY)) write.csv(CoolingTower_Sites, 'N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv', row.names = F)
paste3 <- function(...,sep=" ") { L <- list(...) L <- lapply(L,function(x) {x[is.na(x)] <- ""; x}) ret <-gsub(paste0("(^",sep,"|",sep,"$)"),"", gsub(paste0(sep,sep),sep, do.call(paste,c(L,list(sep=sep))))) is.na(ret) <- ret=="" ret } EMIS <- read.csv('N:/ABR/TradeWasteCustomers/15062017_ActiveInactive.csv', stringsAsFactors=FALSE, skip=6) %>% select(AG_CONSUMERNUMBER, textbox11, textbox65, textbox43, textbox46, textbox49, textbox53, textbox56) %>% mutate(MASTERID = as.numeric(substr(AG_CONSUMERNUMBER, 3, 9))) EMIS$ADDRESS_STRING <- paste3(paste3(EMIS$textbox11, EMIS$textbox65, EMIS$textbox43), paste3(EMIS$textbox46, EMIS$textbox49), paste3(EMIS$textbox53, "VIC", EMIS$textbox56), sep = ", ") write.csv(EMIS, "N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_unlabelled_EMIS.csv", row.names = F)
Gentrack <- read.csv('file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/Metersequencesdetail_June17.csv', stringsAsFactors=FALSE, skip=3) unlabelled_gentrack <- Gentrack %>% select(INSTALL, INSTALL_ADDRESS) %>% distinct() %>% mutate(MASTERID = as.integer(substr(as.character(INSTALL), 3, 9))) %>% filter(MASTERID %in% EMIS$MASTERID) %>% mutate(MASTERID = as.character(MASTERID)) write.csv(unlabelled_gentrack, "N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_unlabelled_Gentrack.csv", row.names = F)
```{python DHHS parserator} import re import GNAF import itertools import operator
import pandas as pd
unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv", header=None)
site = unlabelled[0]
unlabelled = unlabelled[4].apply(lambda x: re.sub(r', Australia$', "", x))
def accumulate(l): it = itertools.groupby(l, operator.itemgetter(1)) for key, subiter in it: yield key, ' '.join(item[0] for item in subiter)
labelled = unlabelled.apply(lambda x: GNAF.parse(x))
labelled = labelled.apply(lambda x: list(accumulate(x)))
cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX', 'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME', 'POSTCODE', 'STATE_ABBREVIATION']
df = pd.DataFrame(columns=cols)
df['SITE'] = site
df['ADDRESS_STRING'] = unlabelled
for i, row in enumerate(labelled): for item in row: df[item[0]][i] = item[1]
df2 = df.ix[1:]
df2.to_csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_labelled_DHHS.csv', index=False)
```{python EMIS parserator} import re import GNAF import itertools import operator import pandas as pd unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_unlabelled_EMIS.csv", header=None) site = unlabelled[0] unlabelled = unlabelled[8].apply(lambda x: re.sub(r', Australia$', "", x)) def accumulate(l): it = itertools.groupby(l, operator.itemgetter(1)) for key, subiter in it: yield key, ' '.join(item[0] for item in subiter) labelled = unlabelled.apply(lambda x: GNAF.parse(x)) labelled = labelled.apply(lambda x: list(accumulate(x))) cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX', 'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME', 'POSTCODE', 'STATE_ABBREVIATION'] df = pd.DataFrame(columns=cols) df['SITE'] = site df['ADDRESS_STRING'] = unlabelled for i, row in enumerate(labelled): for item in row: df[item[0]][i] = item[1] df2 = df.ix[1:] df2.to_csv('N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_labelled_EMIS.csv', index=False)
```{python Gentrack parserator} import re import GNAF import itertools import operator
import pandas as pd
unlabelled = pd.read_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_unlabelled_Gentrack.csv", header=None)
site = unlabelled[2]
unlabelled = unlabelled[1]
def accumulate(l): it = itertools.groupby(l, operator.itemgetter(1)) for key, subiter in it: yield key, ' '.join(item[0] for item in subiter)
labelled = unlabelled.apply(lambda x: GNAF.parse(x))
labelled = labelled.apply(lambda x: list(accumulate(x)))
cols = ['SITE','ADDRESS_STRING', 'BUILDING_NAME', 'LOT_NUMBER_PREFIX', 'LOT_NUMBER', 'LOT_NUMBER_SUFFIX', 'LEVEL_TYPE', 'LEVEL_NUMBER', 'FLAT_TYPE', 'FLAT_NUMBER_PREFIX', 'FLAT_NUMBER', 'FLAT_NUMBER_SUFFIX', 'NUMBER_FIRST', 'NUMBER_FIRST_SUFFIX', 'NUMBER_LAST', 'STREET_NAME', 'STREET_TYPE_CODE', 'LOCALITY_NAME', 'POSTCODE', 'STATE_ABBREVIATION']
df = pd.DataFrame(columns=cols)
df['SITE'] = site
df['ADDRESS_STRING'] = unlabelled
for i, row in enumerate(labelled): for item in row: df[item[0]][i] = item[1]
df2 = df.ix[1:]
df2.to_csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_labelled_Gentrack.csv", index=False)
```r labelled_EMIS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/160517_labelled_EMIS.csv", stringsAsFactors = F) manual_training <- labelled_EMIS %>% filter(BUILDING_NAME != "") %>% sample_frac(0.2) manual_training2 <- labelled_EMIS %>% filter(BUILDING_NAME == "") %>% sample_frac(0.005) manual_training3 <- rbind(manual_training, manual_training2) %>% select(ADDRESS_STRING) write.table(manual_training3, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS.csv", row.names = F, col.names = F) manual_training <- labelled_EMIS %>% filter(BUILDING_NAME == "LOT" | FLAT_TYPE == "PREV" | grepl("SH", BUILDING_NAME) | grepl("GRN", BUILDING_NAME) | grepl("GRD", BUILDING_NAME) | grepl("LOWERGROUND", BUILDING_NAME) | grepl("GROUND", FLAT_TYPE) | FLAT_TYPE == "GR" | FLAT_TYPE == "G") %>% select(ADDRESS_STRING) write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS2.csv", row.names = F, col.names = F) manual_training <- labelled_EMIS %>% filter(BUILDING_NAME == "PRINCES HIGHWAY," | BUILDING_NAME == "WELLINGTON PARADE," | BUILDING_NAME == "OLYMPIC BOULEVARD," | BUILDING_NAME == "PARK CRESCENT," | BUILDING_NAME == "MELBOURNE PLACE,") %>% select(ADDRESS_STRING) write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/170517_unlabelled_EMIS3.csv", row.names = F, col.names = F) manual_training <- labelled_EMIS %>% filter(grepl("^[A-Z]{1}\\s.*", STREET_NAME)) %>% select(ADDRESS_STRING) write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/090617_unlabelled_EMIS4.csv", row.names = F, col.names = F) manual_training <- labelled_EMIS %>% filter(grepl("MELBOURNE UNIVERSITY", ADDRESS_STRING)) %>% select(ADDRESS_STRING) write.table(manual_training, "C:/Users/farrelp1/Documents/GNAFAPI/parserator/data/200617_unlabelled_EMIS5.csv", row.names = F, col.names = F)
ABR <- odbcDriverConnect(connection="Driver={SQL Server Native Client 11.0}; server= WVdb1devsql; database=ABR; uid=CWW\\farrelp1; Trusted_Connection=Yes;") GNAF_address <- sqlQuery(ABR, "SELECT DISTINCT [ADDRESS_DETAIL_PID] ,[BUILDING_NAME] ,CASE WHEN [NUMBER_FIRST] % 2 = 0 THEN REPLACE(SUBSTRING([STREET_NAME], 1, 1) + 'EVEN', ' ', '') ELSE REPLACE(SUBSTRING([STREET_NAME], 1, 1) + 'ODD', ' ', '') END AS [KEY] ,[LOT_NUMBER] ,[NUMBER_FIRST] ,[NUMBER_LAST] ,[STREET_NAME] ,[STREET_TYPE_CODE] ,CASE WHEN [STREET_TYPE_CODE] IS NULL THEN [STREET_NAME] ELSE [STREET_NAME] + ' ' + [STREET_TYPE_CODE] END AS [STREET] ,[LOCALITY_NAME] ,[POSTCODE] ,[LONGITUDE] ,[LATITUDE] ,CASE WHEN [NUMBER_LAST] IS NULL THEN 1 ELSE 2 END AS ADDRESS_TYPE ,CONFIDENCE ,[PRIMARY_SECONDARY] FROM [ABR].[dbo].[ADDRESS_VIEW] WHERE [PRIMARY_SECONDARY] = 'P' OR [PRIMARY_SECONDARY] IS NULL;", stringsAsFactors=FALSE)
text_clean <- function(x, field.type = "text") { if(x == "" | is.na(x)) { return(NA) } else if (field.type == "text"){ x <- gsub(",$","",x) return(toupper(x)) } else if (field.type == "num"){ return(as.numeric(gsub(",","",x))) } } text_clean <- Vectorize(text_clean) labelled_DHHS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_labelled_DHHS.csv", stringsAsFactors = F) DHHS_sites <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_unlabelled_DHHS.csv", stringsAsFactors = F) indx <- !is.na(labelled_DHHS$NUMBER_FIRST_SUFFIX) & is.numeric(labelled_DHHS$NUMBER_FIRST_SUFFIX) labelled_DHHS[indx,]$NUMBER_FIRST <- labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX <- NA indx <- grepl("[A-Z]$", labelled_DHHS$NUMBER_FIRST) numFirst <- gsub("([0-9]+)[A-Z]", "\\1", labelled_DHHS[indx,]$NUMBER_FIRST) suffix <- gsub("([0-9]+)([A-Z])", "\\2", labelled_DHHS[indx,]$NUMBER_FIRST) labelled_DHHS[indx,]$NUMBER_FIRST <- numFirst labelled_DHHS[indx,]$NUMBER_FIRST_SUFFIX <- suffix labelled_DHHS <- labelled_DHHS %>% mutate(BUILDING_NAME = text_clean(BUILDING_NAME), LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")), LEVEL_TYPE = text_clean(LEVEL_TYPE), LEVEL_NUMBER = text_clean(LEVEL_NUMBER, field.type = "num"), FLAT_TYPE = text_clean(FLAT_TYPE), FLAT_NUMBER_PREFIX = text_clean(FLAT_NUMBER_PREFIX), FLAT_NUMBER = text_clean(FLAT_NUMBER, field.type = "num"), NUMBER_FIRST = text_clean(NUMBER_FIRST, field.type = "num"), NUMBER_LAST = text_clean(NUMBER_LAST, field.type = "num"), STREET_NAME = text_clean(STREET_NAME), STREET_TYPE_CODE = text_clean(STREET_TYPE_CODE), LOCALITY_NAME = text_clean(LOCALITY_NAME), POSTCODE = text_clean(POSTCODE, field.type = "num"), STATE_ABBREVIATION = text_clean(STATE_ABBREVIATION), KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3), substr(gsub("'","", STREET_NAME), 1,1), ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))), ID = row_number(), ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2)) %>% left_join(select(DHHS_sites, -ADDRESS_STRING), by = "SITE") GNAF_address <- GNAF_address %>% mutate(PID = as.numeric(gsub("GAVIC", "", ADDRESS_DETAIL_PID))) Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.4.cpp") #Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.1.cpp") #Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev7.cpp") #Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev9.cpp") #Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev10.cpp") Clean2 <- function(nonblanks, ID) { # GNAF_address <- data.table::fread("C:/Users/farrelp1/Documents/GNAFAPI/data/May17_GNAF.csv", stringsAsFactors = F) # # colnames(GNAF_address) <- colnames(data.table::fread("C:/Users/farrelp1/Documents/AddressR/data/May17_GNAF_address.csv", stringsAsFactors = F))[1:15] options <- lapply(nonblanks$KEY, function(x) which(x == GNAF_address$KEY)) names(options) <- ID totalscore <- lapply(seq_along(options), function(i) parallelAddressTotalScore6(nonblanks[i,], GNAF_address[options[[i]],])) names(totalscore) <- ID options2 <- lapply(seq_along(options), function(i) cbind(option = suppressWarnings(options[[i]][which(totalscore[[i]] == min(totalscore[[i]]))]), score = suppressWarnings(totalscore[[i]][which(totalscore[[i]] == min(totalscore[[i]]))]))) names(options2) <- names(options) options3 <- options2[unname(unlist(lapply(options2, function(x) nrow(x) > 0)))] finalresult <- plyr::ldply(names(options3), function(i) cbind(id = i,options3[[i]], GNAF_address[options2[[i]][,1],])) # finalresult <- plyr::ldply(names(options3), function(i) cbind(id = i,options3[[i]], GNAF_address[options2[[i]][,1],])) %>% # mutate(id = as.numeric(as.character(id))) %>% # group_by(id) %>% # filter(CONFIDENCE == max(CONFIDENCE)) #finalresult <- plyr::ldply(options4, function(x) x %>% filter(CONFIDENCE == max(CONFIDENCE))) #%>% #mutate(ID = as.numeric(.id)) #dup <- finalresult[duplicated(finalresult$.id),] #CWW_TradeWaste_ABR <- select(finalresult, AG_CONSUMERNUMBER, ADDRESS_DETAIL_PID) %>% distinct() return(finalresult) } Clean3 <- function(nonblanks, ID) { GNAF_address2 <- data.table::as.data.table(GNAF_address) setkey(GNAF_address2, "KEY") nonblanks2 <- data.table::as.data.table(nonblanks) setkey(nonblanks2, "KEY") totalscore2 <- lapply(1:nrow(nonblanks2), function(i) { if(!is.na(nonblanks2[[i, "NUMBER_FIRST"]])) { options <- GNAF_address2[GNAF_address2[["KEY"]] == nonblanks2[[i, "KEY"]] & GNAF_address2[["NUMBER_FIRST"]] < nonblanks2[[i, "NUMBER_FIRST"]] + 50 & GNAF_address2[["NUMBER_FIRST"]] > nonblanks2[[i, "NUMBER_FIRST"]] - 50] } else if(!is.na(nonblanks2[[i, "LOT_NUMBER"]])) { options <- GNAF_address2 %>% filter(!is.na(LOT_NUMBER)) } else { # options <- GNAF_address2[gsub("([A-Z])(.*)", "\\1", GNAF_address2[["KEY"]]) == gsub("([A-Z])(.*)", "\\1", nonblanks2[[i, "KEY"]])] options <- GNAF_address2 %>% filter(!is.na(BUILDING_NAME)) } result <- parallelAddressTotalScore6(nonblanks2[i,], options) colnames(result) = options[["ADDRESS_DETAIL_PID"]] return(result) }) names(totalscore2) <- nonblanks2[["SITE"]] setkey(GNAF_address2, "ADDRESS_DETAIL_PID") finalresult <- plyr::ldply(names(totalscore2), function(i) { if(length(totalscore2[[i]]) > 0) { ADDRESS_DETAIL_PID <- names(totalscore2[[i]][,which.min(totalscore2[[i]])]) score <- as.numeric(unname(totalscore2[[i]][,which.min(totalscore2[[i]])])) result <- as.data.table(cbind(id = as.numeric(i), ADDRESS_DETAIL_PID = ADDRESS_DETAIL_PID, score = score)) setkey(result, "ADDRESS_DETAIL_PID") return(GNAF_address2[result]) } else{} }) finalresult$score <- as.numeric(finalresult$score) return(finalresult) } #AddressR::AddressClean2 #DHHS_GNAF <- AddressR::AddressClean2(nonblanks=labelled_DHHS, ID=labelled_DHHS$ID) DHHS_GNAF <- Clean3(nonblanks=labelled_DHHS, ID=labelled_DHHS$SITE) # for (i in seq_along(options)) { # # r <- parallelAddressTotalScore3(nonblanks[i, ], GNAF_address[options[[i]], ]) # # } # test <- parallelAddressTotalScore9(labelled_DHHS[1,], GNAF_address) works but very slow #test <- parallelAddressTotalScore10(labelled_DHHS[1:2,], GNAF_address)
text_clean <- function(x, field.type = "text") { if(x == "" | is.na(x)) { return(NA) } else if (field.type == "text"){ x <- gsub(",$","",x) return(toupper(x)) } else if (field.type == "num"){ return(as.numeric(gsub(",","",x))) } } text_clean <- Vectorize(text_clean) EMIS_sites <- read.csv('N:/ABR/TradeWasteCustomers/15062017_ActiveInactive.csv', stringsAsFactors=FALSE, skip=6) %>% select(AG_CONSUMERNUMBER, AG_COMPANYNAME, AG_TRADINGNAME, textbox11, textbox65, textbox43, textbox46, textbox49, textbox53, textbox56, PLD_PROPERTYXYCOORDINATE) %>% mutate(AG_COMPANYNAME = ifelse(AG_TRADINGNAME == "", AG_COMPANYNAME, AG_TRADINGNAME), SITE = AG_CONSUMERNUMBER, LAT=as.integer(gsub("([0-9]{6})([0-9]{7})","\\1",PLD_PROPERTYXYCOORDINATE)), LON=as.integer(gsub("([0-9]{6})([0-9]{7})","\\2",PLD_PROPERTYXYCOORDINATE))) CWWBoundary <- rgdal::readOGR("N:/Asset Information/MUNSYS MapInfo Data/Production/Data/CWW Boundary_2014_region.shp", layer="CWW Boundary_2014_region") EMIS_sp <- EMIS_sites %>% filter(!is.na(LAT) & LAT != 0 & LAT != 999999) EMIS_sp <- sp::SpatialPointsDataFrame(coords=data.frame(X=EMIS_sp$LAT, Y=EMIS_sp$LON), data=data.frame(SITE=EMIS_sp$SITE), proj4string=sp::CRS(proj4string(CWWBoundary))) EMIS_sp <- sp::spTransform(EMIS_sp, sp::CRS("+init=epsg:4326 +proj=longlat")) EMIS_sites <- EMIS_sites %>% left_join(data.frame(SITE=EMIS_sp@data, LATITUDE = EMIS_sp@coords[,2], LONGITUDE = EMIS_sp@coords[,1])) labelled_EMIS <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_labelled_EMIS.csv", stringsAsFactors = F) %>% #left_join(select(EMIS_sites, SITE, AG_COMPANYNAME)) %>% mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>% mutate(#BUILDING_NAME = ifelse(BUILDING_NAME == "", AG_COMPANYNAME, BUILDING_NAME), MASTERID = as.integer(substr(MASTERID, 3, 9))) indx <- !is.na(labelled_EMIS$NUMBER_FIRST_SUFFIX) & is.numeric(labelled_EMIS$NUMBER_FIRST_SUFFIX) if(any(indx == TRUE)) { labelled_EMIS[indx,]$NUMBER_FIRST <- labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- NA } indx <- grepl("[A-Z]$", labelled_EMIS$NUMBER_FIRST) if(any(indx == TRUE)) { numFirst <- gsub("([0-9]+)[A-Z]", "\\1", labelled_EMIS[indx,]$NUMBER_FIRST) suffix <- gsub("([0-9]+)([A-Z])", "\\2", labelled_EMIS[indx,]$NUMBER_FIRST) labelled_EMIS[indx,]$NUMBER_FIRST <- numFirst labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- suffix } indx <- grepl("^[A-Z]{1}\\s.*", labelled_EMIS$STREET_NAME) if(any(indx == TRUE)) { suffix <- gsub("^([A-Z]{1})\\s(.*)", "\\1", labelled_EMIS[indx,]$STREET_NAME) streetname <- gsub("^([A-Z]{1})\\s(.*)", "\\2", labelled_EMIS[indx,]$STREET_NAME) labelled_EMIS[indx,]$NUMBER_FIRST_SUFFIX <- suffix labelled_EMIS[indx,]$STREET_NAME <- streetname } indx <- grepl("^MOUNT\\s.*", labelled_EMIS$STREET_NAME) labelled_EMIS$STREET_NAME[indx] <- gsub("^(MOUNT)", "MT", labelled_EMIS[indx,]$STREET_NAME) indx <- grepl("^17C\\s.*", labelled_EMIS$STREET_NAME) labelled_EMIS$STREET_NAME[indx] <- gsub("^(17C\\s)(.*)", "\\2", labelled_EMIS[indx,]$STREET_NAME) labelled_EMIS <- labelled_EMIS %>% mutate(BUILDING_NAME = unname(text_clean(BUILDING_NAME)), LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")), LEVEL_TYPE = unname(text_clean(LEVEL_TYPE)), LEVEL_NUMBER = unname(text_clean(LEVEL_NUMBER, field.type = "num")), FLAT_TYPE = unname(text_clean(FLAT_TYPE)), FLAT_NUMBER_PREFIX = unname(text_clean(FLAT_NUMBER_PREFIX)), FLAT_NUMBER = unname(text_clean(FLAT_NUMBER, field.type = "num")), NUMBER_FIRST = unname(text_clean(NUMBER_FIRST, field.type = "num")), NUMBER_LAST = unname(text_clean(NUMBER_LAST, field.type = "num")), STREET_NAME = unname(text_clean(STREET_NAME)), STREET_TYPE_CODE = unname(text_clean(STREET_TYPE_CODE)), LOCALITY_NAME = unname(text_clean(LOCALITY_NAME)), POSTCODE = unname(text_clean(POSTCODE, field.type = "num")), STATE_ABBREVIATION = unname(text_clean(STATE_ABBREVIATION)), KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3), substr(gsub("'","", STREET_NAME), 1,1), ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))), #ID = row_number(), ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2)) #%>% #left_join(labelled_Gentrack, by = "MASTERID") labelled_Gentrack <- read.csv("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/8. Gentrack/190617_labelled_Gentrack.csv", stringsAsFactors = F) %>% mutate(MASTERID = SITE) labelled_Gentrack <- labelled_Gentrack %>% mutate(BUILDING_NAME = unname(text_clean(BUILDING_NAME)), LOT_NUMBER = unname(text_clean(LOT_NUMBER, field.type = "num")), LEVEL_TYPE = unname(text_clean(LEVEL_TYPE)), LEVEL_NUMBER = unname(text_clean(LEVEL_NUMBER, field.type = "num")), FLAT_TYPE = unname(text_clean(FLAT_TYPE)), FLAT_NUMBER_PREFIX = unname(text_clean(FLAT_NUMBER_PREFIX)), FLAT_NUMBER = unname(text_clean(FLAT_NUMBER, field.type = "num")), NUMBER_FIRST = unname(text_clean(NUMBER_FIRST, field.type = "num")), NUMBER_LAST = unname(text_clean(NUMBER_LAST, field.type = "num")), STREET_NAME = unname(text_clean(STREET_NAME)), STREET_TYPE_CODE = unname(text_clean(STREET_TYPE_CODE)), LOCALITY_NAME = unname(text_clean(LOCALITY_NAME)), POSTCODE = unname(text_clean(POSTCODE, field.type = "num")), STATE_ABBREVIATION = unname(text_clean(STATE_ABBREVIATION)), KEY = gsub(" ", "", paste0(#substr(LOCALITY_NAME, 1,3), substr(gsub("'","", STREET_NAME), 1,1), ifelse(NUMBER_FIRST %% 2 == 0, "EVEN", "ODD"))), #ID = row_number(), ADDRESS_TYPE = ifelse(is.na(NUMBER_LAST), 1, 2)) # labelled_Gentrack <- labelled_Gentrack %>% # left_join(select(mutate(Gentrack, # MASTERID = as.integer(substr(as.character(INSTALL), 3, 9))), # MASTERID, # AG_COMPANYNAME)) %>% # #mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>% # mutate(BUILDING_NAME = ifelse(is.na(BUILDING_NAME), AG_COMPANYNAME, BUILDING_NAME)) labelled_EMIS2 <- labelled_EMIS lapply(1:nrow(labelled_EMIS), function(i) { NUMBER_FIRST = labelled_EMIS[i,]$NUMBER_FIRST LOT_NUMBER = labelled_EMIS[i,]$LOT_NUMBER MASTERID = labelled_EMIS[i,]$MASTERID REPLACEMENT = labelled_Gentrack[labelled_Gentrack$MASTERID == MASTERID,] #%>% #filter(!is.na(NUMBER_FIRST)) %>% #distinct() if(nrow(REPLACEMENT) > 0) { if(is.na(NUMBER_FIRST) & is.na(LOT_NUMBER)) { labelled_EMIS2 <<- rbind(labelled_EMIS2, REPLACEMENT) labelled_EMIS2 <<- labelled_EMIS2[-i,] } } }) labelled_EMIS2 <- labelled_EMIS2 %>% left_join(select(EMIS_sites, SITE, AG_COMPANYNAME, LATITUDE, LONGITUDE)) %>% #mutate(MASTERID = as.character(format(SITE, scientific=FALSE))) %>% mutate(BUILDING_NAME = ifelse(is.na(BUILDING_NAME), AG_COMPANYNAME, BUILDING_NAME)) %>% distinct()
#Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.cpp") Rcpp::sourceCpp("C:/Users/farrelp1/Documents/AddressR/src/AddressMatchTotalScorev6.4.cpp") # labelled_EMIS2$LATITUDE = rep(as.numeric(NA), length=nrow(labelled_EMIS2)) # # labelled_EMIS2$LONGITUDE = rep(as.numeric(NA), length=nrow(labelled_EMIS2)) # sample <- labelled_EMIS %>% mutate(SITE=as.numeric(SITE)) %>% filter(!is.na(BUILDING_NAME) & is.na(NUMBER_FIRST)) %>% sample_n(10) EMIS_GNAF <- Clean3(nonblanks=labelled_EMIS2, ID=labelled_EMIS2 %>% distinct() %>% .[["SITE"]])
DHHS_GNAF2 <- DHHS_GNAF %>% select(ADDRESS_DETAIL_PID, id, NUMBER_FIRST:LATITUDE) %>% mutate(id = as.numeric(id)) %>% #left_join(CoolingTower_Sites, by=c("id" = "SITE")) %>% left_join(select(labelled_DHHS, SITE, BUILDING_NAME), by=c("id" = "SITE")) %>% as.data.table() setkey(DHHS_GNAF2, "ADDRESS_DETAIL_PID") #EMIS_GNAF2 <- as.data.table(EMIS_GNAF) EMIS_GNAF2 <- EMIS_GNAF %>% select(ADDRESS_DETAIL_PID, id) %>% mutate(id = as.numeric(id)) %>% #left_join(CoolingTower_Sites, by=c("id" = "SITE")) %>% #left_join(select(labelled_EMIS, SITE, BUILDING_NAME), by=c("id" = "SITE")) %>% as.data.table() setkey(EMIS_GNAF2, "ADDRESS_DETAIL_PID") overlay <- EMIS_GNAF2[DHHS_GNAF2]
# options <- EMIS_GNAF[EMIS_GNAF$STREET == overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$STREET,] # # options$EVEN <- ifelse(options$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD") # # overlay$EVEN <- ifelse(overlay$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD") # # match <- options[overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$NUMBER_FIRST <= options$NUMBER_LAST & overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$NUMBER_FIRST >= options$NUMBER_FIRST & overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$EVEN == options$EVEN,]$ADDRESS_DETAIL_PID # # match <- unique(match[!is.na(match)]) # # names(match) <- rep(overlay[overlay$ADDRESS_DETAIL_PID == "GAVIC411790346"]$i.id, length(match)) # # match <- plyr::ldply(match, function(x) c(ADDRESS_DETAIL_PID = unname(x), SITE = names(x))) %>% # left_join(EMIS_GNAF) %>% # .[["id"]] overlay$EVEN <- ifelse(overlay$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD") range_check <- function(ID, STREET, NUMBER_FIRST, EVEN, EMIS_GNAF) { if(is.na(ID)) { options <- EMIS_GNAF[EMIS_GNAF$STREET == STREET,] if(nrow(options) > 0) { options$EVEN <- ifelse(options$NUMBER_FIRST %% 2 == 0, "EVEN", "ODD") match <- options[NUMBER_FIRST <= options$NUMBER_LAST & NUMBER_FIRST >= options$NUMBER_FIRST & EVEN == options$EVEN,]$ADDRESS_DETAIL_PID match2 <- unique(match[!is.na(match)]) #names(match2) <- rep("ADDRESS_DETAIL_PID", length(match2)) if(length(match2) > 0) { match3 <- plyr::ldply(match2) %>% left_join(EMIS_GNAF, by = c("V1" = "ADDRESS_DETAIL_PID")) %>% .[["id"]] return(match3) } } } } results <- sapply(1:nrow(overlay), function(x) range_check(overlay$id[x], overlay$STREET[x], overlay$NUMBER_FIRST[x], overlay$EVEN[x], EMIS_GNAF)) names(results) <- overlay$i.id # df <- data.frame(matrix(unlist(results), nrow=length(unlist(results[unlist(lapply(results, function(x) length(x) > 0))])), byrow=T)) df <- plyr::ldply(lapply(seq_along(results), function(i) data.frame(CONSUMER = results[[i]], i.id = rep(names(results[i]), length(results[[i]])), stringsAsFactors = F))) %>% mutate(i.id = as.numeric(i.id)) #do( data.frame(column = rep(results, each = length(results)), stringsAsFactors = FALSE) ) overlay <- overlay %>% left_join(df) %>% mutate(id = ifelse(is.na(id), CONSUMER, id)) %>% select(ADDRESS_DETAIL_PID:BUILDING_NAME) %>% filter(is.na(id)) %>% distinct()
# VICMAP_LandUse <- readOGR("N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/2. Zones/ll_gda94/plan_overlay.shp", # layer="plan_overlay") gap <- sp::SpatialPointsDataFrame(coords = data.frame(X=overlay$LONGITUDE, Y=overlay$LATITUDE), data=data.frame(id = overlay$i.id), proj4string=sp::CRS("+init=epsg:4326 +proj=longlat")) %>% sp::spTransform(sp::CRS(sp::proj4string(CWWBoundary))) #VICMAP_LandUse2 <- spTransform(VICMAP_LandUse, CRS(proj4string(CWWBoundary))) #proj4string(CWWBoundary) gap_cww <- gap[which(rgeos::gContains(CWWBoundary, gap, byid = TRUE)), ] gap_cww <- gap_cww@data %>% left_join(select(overlay, 1, i.id:BUILDING_NAME), by = c("id"= "i.id"))
# STREET_SCORE <- lapply(gap_cww, function(x) { # if(names(x) == "STREET") { # STREET_SCORE <- stringdist(x$STREET, EMIS_GNAF$STREET) # } # #SUBURB_SCORE <- stringdist(x$LOCALITY_NAME, EMIS_GNAF$LOCALITY_NAME) # }) # # # idx <- lapply(gap_cww, is.character) # # gap_cww[,unname(unlist(idx))][,-1] # # idx <- lapply(EMIS_GNAF, is.character) # # EMIS_GNAF[,unname(unlist(idx))][,!c("KEY", "PRIMARY_SECONDARY", "id")] gap <- data.table(gap_cww) add <- function(x) Reduce("+", x) subtract <- function(x) Reduce("-", x) EMIS2 <- data.table(EMIS_GNAF) cnames <- c("BUILDING_NAME", "STREET", "LOCALITY_NAME") score <- list() for(cname in cnames) { score[[cname]] <- stringdistmatrix(gap[[cname]], EMIS2[[cname]]) } result <- matrix(nrow = nrow(gap), ncol = nrow(EMIS2)) numdiff <- function(x, y) {abs(x - y)} #test <- Map(numdiff, gap[["NUMBER_FIRST"]], EMIS2[["NUMBER_FIRST"]]) #x <- foreach(i=gap[["NUMBER_FIRST"]]) %do% numdiff(i, EMIS2[["NUMBER_FIRST"]]) #x <- foreach(i=EMIS2[["NUMBER_FIRST"]], .combine='rbind') %do% numdiff(i, gap[["NUMBER_FIRST"]]) cnames <- c("NUMBER_FIRST", "NUMBER_LAST") for(cname in cnames) { score[[cname]] <- foreach(i=EMIS2[[cname]], .combine='cbind') %do% numdiff(i, gap[[cname]]) dimnames(score[[cname]]) <- NULL } #apply(gap, c(1,2), function(x) ) # for(cname in cnames) { # # for(i in 1:nrow(gap)) { # # for(j in 1:nrow(EMIS2)) # # #if(!is.na(gap[[cname]][i])) # # result[i,j] <- ifelse(is.na(abs(gap[[cname]][i] - EMIS2[[cname]][j])), 0, abs(gap[[cname]][i] - EMIS2[[cname]][j])) # # } # # score[[cname]] <- result # # } #subtract(list(gap[["NUMBER_FIRST"]], EMIS2[["NUMBER_FIRST"]])) score2 <- add(score) score3 <- apply(score2, 1, which.min) closest_match <- data.table(gap[["id"]], EMIS2[score3,])
write.csv(EMIS_GNAF, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/7. EMIS/150617_EMIS_GNAF.csv", row.names = F) write.csv(DHHS_GNAF, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_DHHS_GNAF.csv", row.names = F) write.csv(gap_cww, "file:///N:/ABR/Output 1.1 New Trade Waste Customers for 2016-17/5. Cooling Towers/160517_DHHS_gap.csv", row.names = F)
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.