1 00:00:00,000 --> 00:00:04,959 I just wanted to make a short note about the data set that I use in the regular 2 00:00:04,959 --> 00:00:08,126 expression lecture about grep, regex or, and sub and Gsub. 3 00:00:08,126 --> 00:00:13,026 So in that lecture, I use a specific data set, which is the Baltimore City Homicide 4 00:00:13,205 --> 00:00:16,910 Event database. You can look at this data in map form by 5 00:00:16,910 --> 00:00:22,347 going to this website up here and I'll put the link in Coursera so that you can find 6 00:00:22,347 --> 00:00:25,933 it later. But basically the, the website is hosted 7 00:00:25,933 --> 00:00:31,072 by the Baltimore Sun the city's newspaper, and they've kept track of every homicide 8 00:00:31,072 --> 00:00:37,560 in the city that's occurred since 2007 up to the current, the present day. 9 00:00:37,560 --> 00:00:43,598 So, for example, you can see here that there have been 171 homicides in 2012 in 10 00:00:43,598 --> 00:00:46,780 Baltimore City. So the map is interactive. 11 00:00:46,780 --> 00:00:52,431 You can do things like, you know, select on let's say, the past let's say, the past 12 00:00:52,431 --> 00:00:55,322 six months. You can select a district of Baltimore if 13 00:00:55,322 --> 00:00:57,955 you want or you can just look at all the districts. 14 00:00:57,955 --> 00:01:00,950 A zip code an age of the victim, a gender of the victim, 15 00:01:01,105 --> 00:01:04,564 And then the race of the victim, As well as the cause of death here. 16 00:01:04,564 --> 00:01:09,090 There's a few different causes of death. So, you can show the results and I'm just 17 00:01:09,090 --> 00:01:13,121 picking all the homicides that have occurred in the last six months. And then, 18 00:01:13,121 --> 00:01:17,570 you'll get a map, a little Google map like this that'll show you where each victim 19 00:01:17,570 --> 00:01:21,444 was found in the city and they're color coded by the cause of death. 20 00:01:21,444 --> 00:01:25,632 So the red, you can see that the vast majority of the homicides here in the last 21 00:01:25,632 --> 00:01:28,406 six months were red and that indicates a shooting. 22 00:01:28,563 --> 00:01:32,804 And then, if you zoom in a little bit, let's say, if you zoom in on this portion 23 00:01:32,804 --> 00:01:36,323 right here. You can see, the, you can click on one of 24 00:01:36,323 --> 00:01:41,668 the markers and you'll see the name of the person the street, the street address of 25 00:01:41,668 --> 00:01:46,953 where the person was found, a little bit of information, the race, gender, age, and 26 00:01:46,953 --> 00:01:52,239 then where the person died, and then the cause of death which was determined in 27 00:01:52,239 --> 00:01:56,503 this case to be a shooting. So you can click on the, each marker like 28 00:01:56,503 --> 00:01:59,266 that. You'll get information about each victim 29 00:01:59,446 --> 00:02:05,221 and you can kind of browse the map like this. and then if you scroll down a little 30 00:02:05,221 --> 00:02:10,967 bit more, you'll see that there's a little bit of statistical analysis here, so 31 00:02:10,967 --> 00:02:16,372 basically just a bar chart of the number of homicides that occurred each month. 32 00:02:16,577 --> 00:02:20,340 And then, and they're color coded by the cause of death. 33 00:02:20,340 --> 00:02:25,881 And the little line graph here, is the number of homicides per month in the 34 00:02:25,881 --> 00:02:30,999 previous year. so, so, you can compare this year in the bars graph and the last 35 00:02:30,999 --> 00:02:34,873 year's in the line graph. If you want to look at previous year's 36 00:02:34,873 --> 00:02:38,872 data in this format, you can click on the little arrow over here. 37 00:02:38,872 --> 00:02:44,558 And now, you can look at monthly homicides in 2011, monthly homicides in 2010, etc., 38 00:02:44,558 --> 00:02:47,933 like that. So feel free to take a look at this data. 39 00:02:48,120 --> 00:02:53,570 unfortunately, the data set is not easily downloadable in, in a, in a batch form so 40 00:02:53,570 --> 00:02:57,397 that we can look at it for example, and load it into R. 41 00:02:57,397 --> 00:03:03,209 So, what I have done is essentially cut and paste the data from the HTML source, 42 00:03:03,422 --> 00:03:09,162 and created a text file, which has a, which has lot of the HTML source in there, 43 00:03:09,162 --> 00:03:15,399 but essentially boils down to having a one homicide record per line in the text file. 44 00:03:15,612 --> 00:03:21,140 So, if you want to open, you can open up the text file, which I'll do right here. 45 00:03:21,140 --> 00:03:26,467 I'll, first of all, unzip the file. And if you take a look at it, just in your 46 00:03:26,467 --> 00:03:31,277 the text editor or whatever application you want to use, you'll notice that there 47 00:03:31,277 --> 00:03:34,771 is a lot of text here. But basically, the first line of data 48 00:03:34,771 --> 00:03:36,988 looks like this. So, this is the first line. 49 00:03:36,988 --> 00:03:40,631 It does wrap around a little bit because it's too long for my screen. 50 00:03:40,789 --> 00:03:45,118 But you can see there, there's information about lat, the latitude and longitude of 51 00:03:45,118 --> 00:03:49,289 where the person was found the person's name, street address where they were 52 00:03:49,289 --> 00:03:52,562 found, and some information about their race, gender, and age. 53 00:03:52,720 --> 00:03:55,571 And the date that they were found and where they died. 54 00:03:55,571 --> 00:03:59,531 And this person, in this case, this person died at shock trauma which is the 55 00:03:59,531 --> 00:04:03,859 University of Maryland's emergency room. So, you can see there's one line per 56 00:04:03,859 --> 00:04:06,878 victim. There's about a little over 1200 lines in 57 00:04:06,878 --> 00:04:09,871 this file, So 1200 homicides over the last five 58 00:04:09,871 --> 00:04:12,927 years. And so, you can take a look at this data 59 00:04:12,927 --> 00:04:15,796 set. You can read it into R, or you can view it 60 00:04:15,983 --> 00:04:18,540 on the map provided by the Baltimore Sun.