bbarrilleaux · February 18, 2014 05:07
diff --git a/statepops.csv b/statepops.csv
diff --git a/UFOsightings.R b/UFOsightings.R
 library("maps")
 library("ggplot2")
 library("RColorBrewer")
 library("mapproj")

 # data from infochimps, http://www.infochimps.com/datasets/60000-documented-ufo-sightings-with-text-descriptions-and-metada

 tsvfile <- "chimps_16154-2010-10-20_14-33-35/ufo_awesome.tsv"
 ufo <- read.table(tsvfile, sep ="\t", fill=TRUE, stringsAsFactors = FALSE)

 ufo$state <- sapply(ufo$V3, function (x) strsplit(as.character(x), ", ")[[1]][2])
 ufo$statename <- tolower(state.name[match(ufo$state, state.abb)])

 # this file is given in the gist:
 popfile <- "statepops.csv"
 statePops <- read.csv(popfile, header = FALSE)
 statePops <- statePops[, 3 : 4]
 statePops$region <- tolower(state.name[match(statePops$V3, state.abb)])
 names(statePops)[2] <- "population"

 stateTable <- unlist(table(ufo$statename))
 stateDF <- as.data.frame(stateTable)
 names(stateDF) <- c("region", "UFO")

 stateDF <- merge(stateDF, statePops)
 stateDF$UFOpercapita <- stateDF$UFO * 10000 / stateDF$population

 # set up a theme for the map

 new_theme_empty <- theme_bw()  
 new_theme_empty$line <- element_blank()
 new_theme_empty$rect <- element_blank()
 new_theme_empty$strip.text <- element_blank()
 new_theme_empty$axis.text <- element_blank()
 new_theme_empty$axis.title <- element_blank()
 new_theme_empty$plot.margin <- structure(c(0, 0, -1, -1), unit = "lines", valid.unit = 3L, class = "unit")
 
 stateShapes <- map("state", plot = FALSE, fill = TRUE)
 stateShapes <- fortify(stateShapes)  # Load state shapefiles and convert to a data.frame

 stateShapes$UFO = stateDF$UFOpercapita[match(stateShapes$region, stateDF$region)]  

 # make the map!

 myPalette <- brewer.pal(9,"PuRd")

 mapPlot <- ggplot(stateShapes,
                  aes(x = long, y = lat, group = group,
                      fill = UFO))
 mapPlot <- mapPlot + geom_polygon(colour = "BLACK")
 mapPlot <- mapPlot + coord_map(project="conic", lat0 = 30)
 mapPlot <- mapPlot + new_theme_empty
 mapPlot <- mapPlot + scale_fill_gradientn("UFO sightings per 10,000 residents",
                                          colours = myPalette)
 mapPlot <- mapPlot + ggtitle("UFO sightings, 1995-2010")
 print(mapPlot)

 # make the word cloud!
 library(tm)
 library(wordcloud)
 WAtext <- data.frame(as.character(ufo[ufo$state == "WA", 6]))
 WAtext <- data.frame(WAtext[!is.na(WAtext)])
 WA.corpus <- Corpus(DataframeSource(WAtext))

 WA.corpus <- tm_map(WA.corpus, removePunctuation)
 WA.corpus <- tm_map(WA.corpus, tolower)
 WA.corpus <- tm_map(WA.corpus, function(x) removeWords(x, stopwords("english")))
 tdm <- TermDocumentMatrix(WA.corpus)
 m <- as.matrix(tdm)
 v <- sort(rowSums(m),decreasing=TRUE)
 d <- data.frame(word = names(v),freq=v)
 pal <- brewer.pal(9, "BuGn")
 pal <- pal[-(1:2)]
 wordcloud(d$word, d$freq, scale = c(5, .5), min.freq = 3, max.words = 60, random.order = TRUE, rot.per = 0.15, colors = pal, vfont = c("sans serif", "plain"))
1	California	CA	36756666	403931.96	91	155959	235.68
2	Texas	TX	24326974	678051.12	35.88	261797	92.92
3	New York	NY	19490297	122283.7	159.39	47214	412.81
4	Florida	FL	18328340	139760.29	131.14	53927	339.87
5	Illinois	IL	12901563	143961.9	89.62	55584	232.11
6	Pennsylvania	PA	12448279	116075.5	107.24	44817	277.76
7	Ohio	OH	11485910	106054.83	108.3	40948	280.5
8	Michigan	MI	10003422	147121.68	67.99	56804	176.1
9	Georgia	GA	9685744	149975.85	64.58	57906	167.27
10	North Carolina	NC	9222414	126160.91	73.1	48711	189.33
11	New Jersey	NJ	8682661	19209.94	451.99	7417	1170.64
12	Virginia	VA	7769089	102547.99	75.76	39594	196.22
13	Washington	WA	6549224	172348.17	38	66544	98.42
14	Arizona	AZ	6500180	294313.3	22.09	113635	57.2
15	Massachusetts	MA	6497967	20305.51	320.01	7840	828.82
16	Indiana	IN	6376792	92895.1	68.65	35867	177.79
17	Tennessee	TN	6214888	106751.54	58.22	41217	150.78
18	Missouri	MO	5911605	178413.92	33.13	68886	85.82
19	Maryland	MD	5633597	25314.54	222.54	9774	576.39
20	Wisconsin	WI	5627967	140662.25	40.01	54310	103.63
21	Minnesota	MN	5220393	206188.95	25.32	79610	65.57
22	Colorado	CO	4939456	268628.39	18.39	103718	47.62
23	Alabama	AL	4661900	131426.36	35.47	50744	91.87
24	South Carolina	SC	4479800	77981.95	57.45	30109	148.79
25	Louisiana	LA	4410796	112825.06	39.09	43562	101.25
26	Kentucky	KY	4269245	102895.05	41.49	39728	107.46
27	Oregon	OR	3790060	268631.09	14.11	95997	39.48
28	Oklahoma	OK	3642361	177846.71	20.48	68667	53.04
29	Connecticut	CT	3501252	12548.49	279.02	4845	722.65
30	Iowa	IA	3002555	144700.05	20.75	55869	53.74
31	Mississippi	MS	2938618	121488.57	24.19	46907	62.65
32	Arkansas	AR	2855390	134856	21.17	52068	54.84
33	Kansas	KS	2802134	211899.88	13.22	81815	34.25
34	Utah	UT	2736424	212751.98	12.86	82144	33.31
35	Nevada	NV	2600167	284448.03	9.14	109826	23.68
36	New Mexico	NM	1984356	314310.6	6.31	121356	16.35
37	West Virginia	WV	1814468	62361.73	29.1	24078	75.36
38	Nebraska	NE	1783432	199097.57	8.96	76872	23.2
39	Idaho	ID	1523816	214313.75	7.11	82747	18.42
40	Maine	ME	1316456	79932.21	16.47	30862	42.66
41	New Hampshire	NH	1315809	23227.01	56.65	8968	146.72
42	Hawaii	HI	1288198	16635.49	77.44	6423	200.56
43	Rhode Island	RI	1050788	2706.54	388.24	1045	1005.54
44	Montana	MT	967440	376977.95	2.57	145552	6.65
45	Delaware	DE	873092	5060.84	172.52	1954	446.82
46	South Dakota	SD	804194	196541.25	4.09	75885	10.6
47	Alaska	AK	686293	1481346	0.46	571951	1.2
48	North Dakota	ND	641481	178647.02	3.59	68976	9.3
49	Vermont	VT	621270	23957.39	25.93	9250	67.16
50	Wyoming	WY	532668	251487.85	2.12	97100	5.49
	library("maps")
	library("ggplot2")
	library("RColorBrewer")
	library("mapproj")

	# data from infochimps, http://www.infochimps.com/datasets/60000-documented-ufo-sightings-with-text-descriptions-and-metada

	tsvfile <- "chimps_16154-2010-10-20_14-33-35/ufo_awesome.tsv"
	ufo <- read.table(tsvfile, sep ="\t", fill=TRUE, stringsAsFactors = FALSE)

	ufo$state <- sapply(ufo$V3, function (x) strsplit(as.character(x), ", ")[[1]][2])
	ufo$statename <- tolower(state.name[match(ufo$state, state.abb)])

	# this file is given in the gist:
	popfile <- "statepops.csv"
	statePops <- read.csv(popfile, header = FALSE)
	statePops <- statePops[, 3 : 4]
	statePops$region <- tolower(state.name[match(statePops$V3, state.abb)])
	names(statePops)[2] <- "population"

	stateTable <- unlist(table(ufo$statename))
	stateDF <- as.data.frame(stateTable)
	names(stateDF) <- c("region", "UFO")

	stateDF <- merge(stateDF, statePops)
	stateDF$UFOpercapita <- stateDF$UFO * 10000 / stateDF$population

	# set up a theme for the map

	new_theme_empty <- theme_bw()
	new_theme_empty$line <- element_blank()
	new_theme_empty$rect <- element_blank()
	new_theme_empty$strip.text <- element_blank()
	new_theme_empty$axis.text <- element_blank()
	new_theme_empty$axis.title <- element_blank()
	new_theme_empty$plot.margin <- structure(c(0, 0, -1, -1), unit = "lines", valid.unit = 3L, class = "unit")

	stateShapes <- map("state", plot = FALSE, fill = TRUE)
	stateShapes <- fortify(stateShapes) # Load state shapefiles and convert to a data.frame

	stateShapes$UFO = stateDF$UFOpercapita[match(stateShapes$region, stateDF$region)]

	# make the map!

	myPalette <- brewer.pal(9,"PuRd")

	mapPlot <- ggplot(stateShapes,
	aes(x = long, y = lat, group = group,
	fill = UFO))
	mapPlot <- mapPlot + geom_polygon(colour = "BLACK")
	mapPlot <- mapPlot + coord_map(project="conic", lat0 = 30)
	mapPlot <- mapPlot + new_theme_empty
	mapPlot <- mapPlot + scale_fill_gradientn("UFO sightings per 10,000 residents",
	colours = myPalette)
	mapPlot <- mapPlot + ggtitle("UFO sightings, 1995-2010")
	print(mapPlot)

	# make the word cloud!
	library(tm)
	library(wordcloud)
	WAtext <- data.frame(as.character(ufo[ufo$state == "WA", 6]))
	WAtext <- data.frame(WAtext[!is.na(WAtext)])
	WA.corpus <- Corpus(DataframeSource(WAtext))

	WA.corpus <- tm_map(WA.corpus, removePunctuation)
	WA.corpus <- tm_map(WA.corpus, tolower)
	WA.corpus <- tm_map(WA.corpus, function(x) removeWords(x, stopwords("english")))
	tdm <- TermDocumentMatrix(WA.corpus)
	m <- as.matrix(tdm)
	v <- sort(rowSums(m),decreasing=TRUE)
	d <- data.frame(word = names(v),freq=v)
	pal <- brewer.pal(9, "BuGn")
	pal <- pal[-(1:2)]
	wordcloud(d$word, d$freq, scale = c(5, .5), min.freq = 3, max.words = 60, random.order = TRUE, rot.per = 0.15, colors = pal, vfont = c("sans serif", "plain"))