prenom: prenom

# library(dplyr)
# library(data.table)
# library(ggplot2)
# library(rgeos)
# library(ggmap)
# library(maptools)
# library(stringdist)

# donnees_insee <- rio::import('./data/dpt2017.txt', encoding = 'UTF-8', setclass = 'data.table')
# map_dep <- rgdal::readOGR("./data/departements-20140306-50m-shp/departements-20140306-50m.shp")
# map_dep <- fortify (map_dep, region="code_insee")

# annee2int <- function(x) {
# 	if (length(x) == 1) {
# 		if (x == 'XXXX') {
# 			return(1800)
# 		} else {
# 			return(as.integer(x))
# 		}
# 	}
# 	x[x == 'XXXX'] <- '1800'
# 	return(as.integer(x))
# }
#
# proportionneur <- function(dt = donnees_insee, sexe = 'F', annee = NULL, departement = NULL){
# 	sexeN <- 1L * (toupper(sexe) == 'F')
# 	if (is.null(annee)) {
# 		filtre_annee <- sort(unique(dt$annais))
# 	} else {
# 		filtre_annee <- annee
# 	}
# 	if (is.null(departement)) {
# 		filtre_departement <- sort(unique(dt$dpt))
# 	} else {
# 		filtre_departement <- departement
# 	}
# 	dts <- dt[sexe > sexeN & annais %in% filtre_annee & dpt %in% filtre_departement,.(total = 1.0 * sum(nombre, na.rm = TRUE)), by = .(annais, dpt)]
# 	dts_annee <- dt[sexe > sexeN & annais %in% filtre_annee & dpt %in% filtre_departement,.(total = 1.0 * sum(nombre, na.rm = TRUE)), by = .(annais)]
# 	dts_dpart <- dt[sexe > sexeN & annais %in% filtre_annee & dpt %in% filtre_departement,.(total = 1.0 * sum(nombre, na.rm = TRUE)), by = .(dpt)]
# 	return(list(total = dts,
# 							annee = dts_annee,
# 							dptot = dts_dpart))
# }
#
# selecteur <- function(dt = donnees_insee, sexe = 'F', annee = NULL, deparatement = NULL) {
# 	sexeN <- 1L * (toupper(sexe) == 'F')
# 	if (is.null(annee)) {
# 		filtre_annee <- sort(unique(dt$annais))
# 	} else {
# 		filtre_annee <- annee
# 	}
# 	if (is.null(deparatement)) {
# 		filtre_departement <- sort(unique(dt$dpt))
# 	} else {
# 		filtre_departement <- departement
# 	}
#
# 	dts <- dt[sexe > sexeN & annais %in% filtre_annee & dpt %in% filtre_departement & annais != 'XXXX' & dpt != 'XX']
# 	dts_annee <- dts[,.(nombre = sum(nombre, na.rm = TRUE)), by = .(preusuel, annais)]
# 	dts_departement <- dts[,.(nombre = sum(nombre, na.rm = TRUE)), by = .(preusuel, departement)]
# 	return(list(total = dts,
# 							annee = dts_annee,
# 							dptot = dts_departement))
# }
#
# histoire_prenom <- function(dt = donnees_insee, prenom = 'GINETTE') {
# 	return(dt[preusuel %in% prenom & annais != 'XXXX' & dpt != 'XX'])
# }
#
# graph_annee <- function(dt = histoire_prenom(), sexe = 'F', departement = NULL) {
# 	sexeN <- 1L * (toupper(sexe) == 'F')
# 	titre <- paste0(unique(dt$preusuel), collapse = ', ')
# 	if (!is.null(departement)) {
# 		donnees <- dt[dpt %in% departement & annais != 'XXXX' & dpt != 'XX', .(nombre = sum(nombre,na.rm = TRUE)), by = .(sexe, annais)]
# 		prop <- proportionneur(donnees_insee, sexe, departement = departement)$annee
# 	} else {
# 		donnees <- dt[annais != 'XXXX' & dpt != 'XX', .(nombre = sum(nombre,na.rm = TRUE)), by = .(sexe, annais)]
# 		prop <- prop_defaut$annee
# 	}
# 	# donnees$Sexe <- factor(x = c('Garçon','Fille'))[donnees$sexe]
# 	donnees <- merge(x = donnees, y = prop, by = c('annais'), all.x = TRUE)[, ratio := (1.0 * nombre) / total ]
# 	gg_abs <- ggplot(data = donnees, aes(x = annee2int(annais), y = nombre)) +
# 		geom_bar(stat = 'identity', fill = 'Firebrick') +
# 		xlab('annee') +
# 		ylab('nombre de naissances') +
# 		ggtitle(titre)
# 	gg_rel <- ggplot(data = donnees, aes(x = annee2int(annais), y = ratio)) +
# 		geom_bar(stat = 'identity', fill = 'Firebrick') +
# 		xlab('annee') +
# 		ylab('part des naissances') +
# 		ggtitle(titre)
# 	return(list(abs = gg_abs,
# 							rel = gg_rel))
# }
#
# graph_departement <- function(dt = histoire_prenom(), annee = NULL, sexe = 'F', map = map_dep) {
# 	if (toupper(sexe) != 'F') {.l <- 0} else {.l <- 1}
# 	if (!is.null(annee)) {
# 		donnees <- dt[annais %in% annee & dpt != 'XX' & annais != 'XXXX' & sexe >.l, .(nombre = sum(nombre,na.rm = TRUE)), by = .(dpt)]
# 		prop <- proportionneur(donnees_insee, sexe, annee = annee)$dptot
# 	} else {
# 		donnees <- dt[dpt != 'XX' & annais != 'XXXX' & sexe >.l, .(nombre = sum(nombre,na.rm = TRUE)), by = .(dpt)]
# 		prop <- prop_defaut$dptot
# 	}
# 	# donnees$Sexe <- factor(x = c('Garçon','Fille'))[donnees$sexe]
# 	# browser()
# 	donnees <- merge(x = donnees, y = prop, by = c('dpt'), all.x = TRUE)[, ratio := 100 * nombre / total ]
# 	titre <- paste0(paste0(unique(dt$preusuel), collapse = ', '), ifelse(.l == 1, ' : Filles', ' : Garçons et filles'))
# 	gg_abs <- ggplot(data = donnees) +
# 		geom_map(aes(map_id = dpt, fill = nombre),
# 						 map = map) +
# 		coord_map() +
# 		expand_limits(x=map$long[nchar(map$id)<3],
# 									y=map$lat[nchar(map$id)<3]) +
# 		scale_fill_gradient(name="Nombre", low="LightYellow", high="Firebrick") +
# 		theme_void() +
# 		theme(legend.position = "bottom",
# 					legend.title = element_blank(),
# 					legend.direction = "horizontal") +
# 		ggtitle(titre)
# 	gg_rel <- ggplot(data = donnees) +
# 		geom_map(aes(map_id = dpt, fill = ratio),
# 						 map = map) +
# 		coord_map() +
# 		expand_limits(x = map$long[nchar(map$id)<3],
# 									y = map$lat[nchar(map$id)<3]) +
# 		scale_fill_gradient(name="Nombre", low="LightYellow", high="Firebrick") +
# 		theme_void() +
# 		theme(legend.position = "bottom",
# 					legend.title = element_blank(),
# 					legend.direction = "horizontal") +
# 		ggtitle(titre)
# 	return(list(abs = gg_abs,
# 							rel = gg_rel))
# }
#
# deparseur <- function(texte, pre_dep = 'pre') {
# 	if (pre_dep == 'pre') {
# 		text1 <- gsub('[^[:alpha:],;]','', texte)
# 		text1 <- str_split(text1, '[,;[:blank:]]')[[1]]
# 		text1 <- sort(unique(toupper(text1[nchar(text1) > 0])))
# 	} else {
# 		text1 <- gsub('[^[:digit:]:,;]','', texte)
# 		text1 <- str_split(text1, '[,;[:blank:]]')[[1]]
# 		verif <- grepl(':',text1)
# 		if (any(verif)) {
# 			text2 <- purrr::map(str_split(text1[verif], '[:]', n = 2), ~.x[1]:.x[2]) %>% unlist(.)
# 			text1 <- c(text1[!verif], text2)
# 		}
# 		text1 <- sort(unique(toupper(text1[nchar(text1) > 0])))
# 	}
# 	return(text1)
# }
#
# prop_defaut <- proportionneur(dt = donnees_insee, sexe = 'F')
#
# asciifier <- function(x) {
# 	x %>%
# 		purrr::map_chr(~gsub("[ÀÂÄ]", 'A',.x)) %>%
# 		purrr::map_chr(~gsub("[Æ]", 'AE',.x)) %>%
# 		purrr::map_chr(~gsub("[Ç]", 'C',.x)) %>%
# 		purrr::map_chr(~gsub("[ÉÈÊË]", 'E',.x)) %>%
# 		purrr::map_chr(~gsub("[ÎÏ]", 'I',.x)) %>%
# 		purrr::map_chr(~gsub("[ÔÖ]", 'O',.x)) %>%
# 		purrr::map_chr(~gsub("[ÙÛÜ]", 'U',.x)) %>%
# 		purrr::map_chr(~gsub("[Ÿ]", 'Y',.x))
# }

# graph_annee()
# graph_departement()
# graph_annee(departement = paste0(c(75,77,78,92,93,94,95)))
# graph_departement(annee = paste0(1980:2017))
# graph_annee(dt = histoire_prenom(prenom = c('ELENA','HELENE', 'HELENA')))
# graph_departement(dt = histoire_prenom(prenom = c('ELENA','HELENE', 'HELENA')),annee = paste0(1980:2017))


# stringdist(a = 'GINETTE', b = 'ELENA', method = 'osa')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'lv')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'dl')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'hamming')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'lcs')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'qgram')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'cosine')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'jaccard')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'jw')
# stringdist(a = 'GINETTE', b = 'ELENA', method = 'soundex')

# liste_prenom <- donnees_insee[preusuel != '_PRENOMS_RARES' | preusuel != 'RARES'] %>%
# 	.$preusue %>%
# 	unique(.) %>%
# 	sort(.) %>%
# 	gsub(pattern = '-', replacement = '', x = .) %>%
# 	asciifier(.) %>%
# 	unique(.) %>%
# 	sort(.)

# liste_prenom %>%
# 	.[grepl("[[:cntrl:]]", stringi::stri_enc_toascii(.))] %>%
# 	gsub(pattern = '[A-Z-]', replacement = '', x = .) %>%
# 	stringr::str_split(., '') %>%
# 	unlist(.) %>%
# 	unique(.) %>%
# 	sort(.)

# Compute hierarchical clustering
# res.hc <- liste_prenom %>%
# 	stringdistmatrix(a = ., method = "osa") %>% # Compute dissimilarity matrix
# 	hclust(method = "ward.D2")