NBA 2014-15 Top 50 Scorer Heatmap

nba_top_50_heatmap

## Custom Basketball Reference Scraper Season Scraper
get_bref_player_season_stats <- 
	function(season.end, stat_type = c("Advanced","Totals","Per Minute","Per Game"), 
					 team.totals = F , league = 'NBA'){
	packages <-
		c('rvest','magrittr','dplyr','stringr','tidyr')
	lapply(packages, library, character.only = T)

	bref_team_base <-
		'http://www.basketball-reference.com/leagues/'
	bref_base <-
		'http://www.basketball-reference.com'

	stat_type <-
		match.arg(stat_type,choices = c("Advanced",'Totals','Per Minute',"Per Game"))

	stat_type %<>%
		tolower %>% gsub("\\ ",'_',.)
	url <-
		paste0(bref_team_base,league,'_',season.end,'_',stat_type,'.html')
	css_page <-
		paste0('#',stat_type)

	css_player <-
		'td:nth-child(2) a'

	page <-
		url %>%
		read_html

	tables <-
		page %>%
		html_table(fill = T)

	data <-
		tables[1] %>%
		data.frame %>%
		tbl_df

	url.player <-
		page %>%
		html_nodes(css_player) %>%
		html_attr('href') %>%
		paste0('http://www.basketball-reference.com/',.)

	names(data) %<>%
		tolower

	data %<>%
		dplyr::filter(!rk == 'Rk')

	data %<>%
		dplyr::select(-rk) %>%
		rename(id.team = tm, id.pos = pos)

	if(team.totals == T){
		multi.teams <-
			data %>%
			dplyr::filter(id.team == "TOT")

		one.team <-
			data %>%
			dplyr::filter(!player %in% multi.teams$player)

		data <-
			bind_rows(multi.teams,one.team) %>%
			arrange(player)
	} else{
		data %<>%
			dplyr::filter(!id.team == 'TOT')
	}

	numerics <-
		data %>%
		dplyr::select(-player, -id.pos, -id.team) %>%
		names

	data[,numerics] %<>%
		apply(2, as.numeric)

	season.start <-
		season.end - 1

	id.season <-
		season.start %>%
		paste0('-',season.end %>% substr(3,4))

	data %<>%
		mutate(season.end,
					 id.season,
					 id.pos = data$id.pos %>% substr(1,2),
					 id.name.table = stat_type,
					 url.season = url)

	data$player.hof <-
		data$player %>%
		grepl('\\*',.)

	data$player %<>%
		gsub('\\*','',.)

	if(c('var.20') %in% names(data)){
		data %<>%
			dplyr::select(-var.20, -var.25)
	}

	return(data)

}

data15 <-
	get_bref_player_season_stats(season.end = 2015,stat_type = "Per Game",team.totals = T)

plot.data <-
	data15 %>%
	arrange(desc(pts)) %>%
	slice(1:50) %>%
	select(player,mp:pts, g) %>%
	select(player,pts,fga,mp,trb,drb,orb,everything())

row.names(plot.data) <-
	plot.data$player

plot <-
	plot.data %>%
	select(pts:g) %>%
	d3heatmap(scale = "column",  color = scales::col_quantile("Blues", NULL, 5),
						dendrogram = "row", k_row = 3)

plot ## look here

packages <- 
	c('materializeR', 'htmltools')
# Or beging to style with materializeCSS

#devtools::install_github("timelyportfolio/materializeR")
lapply(packages, library, character.only = T)

plot %<>%
	tagList(
		materialize(),
		tags$h3("NBA 2014-15 Dendogram Heatmap\nTop 50 Players by Points Per Game")
		,.) %>% html_print()