 Okay, so my name's Yang, for I can tell I built, or rather I built an Isochrome map setting recently, and then that's what I'm presenting for I can tell. So just as an overview, an Isochrome map is, Isochrome maps is that you draw contours of equal travel time for a single point, but it's kind of like a topological map, but instead of elevation, your contours are travel time. So, I mean something I found is like, it's kind of easy to kind of imagine it, by 3D, so this is the Isochrome maps projected into 3D, starting from Bishan. So, this is for public transport. Folks have done it for driving before, but public transport in Singapore is particularly difficult because it's just hard to find a clean source of public transport feeds, the GTFS feed. So, what you find are things like this, like little islands or little rich lines on the map itself, because it's kind of places that are faster to commute to or more accessible. So, these are essentially your train stations, right? So, I mean a couple other patterns that we have here. I mean, this is taking from Chachukang, or Bukit Panjan, and the equivalent 3D version of it. So, Singapore is particularly peculiar in this case, but we don't really have any natural barriers except for that thing in the middle, which is the central water catchment area. Like the rest of Singapore is completely fine except for this thing, which kind of makes sense. I mean, okay, yeah, I won't go into that one. So, one of the boys in the office actually lives in this place called Floor Drive. So, for Boston. So, yeah, and so not just, you don't get islands just by, just through MRT stations. You also get islands. These are actually buses that travel down the expressway. So, these are actually, this is Selita Camp, this is somewhere else. Yeah, they are actually buses that kind of just run through this expressway here. Yeah, so, and then on Rated, there was these guys that were looking at Sumai Goodong Camp. So, Sumai Goodong Camp is awesome. It's like no man's land. So, okay, generally speaking, if the gradient is steeper, the travel time is, your speed is slower. If the gradient is gentler, your travel time is actually faster. It's kind of counter-intuitive, but that's not what I suppose. It's just an analogy in this case. So, the thing I like about Isochrome Maps is that it leads to other things. So, it leads to things like this one, which is, oh, where we, here we go, yeah. So, based on that, like, just when this, just when, shortly after I made the Isochrome Maps, the new twas line, actually, the new twas line extension actually came up. So, this is before the extension. So, this is how the maps look like before the extension, starting from Jurong East. So, mainly this area here. This is how it looks like after the extension. So, this area here, right? I mean, particularly this kind, this representation, not particularly great, it's kind of like, spot the difference. So, what I did is actually just conclued the deltas. So, these are the four stations that are included in the extension. And, the brighter the colors, the more time savings up to 25 minutes. So, namely around these two stations, if I'm right, is twas west and twas crescent, or something like that. They have it here, actually. So, it kind of makes sense. Like, if you look at LTA's reported time savings, it makes sense why they report only around twas west or twas crescent, because that's where the most time savings is at. If you look at Gao Circle, and twas link is not as impressive a time saving as the other two, right? So, this is one toy. So, I found this other guy's analysis on accessibility in general in Singapore. Absolutely not in Singapore. In general, you compare jobs that are accessible from a point within, say, X number of minutes. 35, 45, whatever. I try to refer from URA and the best data set for that is CPF, but that's classified. So, instead of using jobs as a measure of accessibility, I made this toy version which measures your accessibility to the nearest your de-hocke center accessibility index. So, within 45 minutes, how many orca centers in Singapore can you get to? So, it's kind of, it's very not really, because up here there's actually a lot of jobs in industrial manufacturing, up down here there's a lot of jobs in manufacturing, but it gives you a sense of accessibility in general. That's it. Just for questions, I see a lot of APIs from LTA for traffic fees, but you said it's normally dirty, can you elaborate? Okay, so to build the Isochronic map what I did, what you need beforehand is, you need this thing called the GTFS, general transit fee specification data set which compiles the public transport, which essentially is what LTA would give to Google for them to do their transport routing. That was back in the day, now they have real-time data as well. From my understanding, LTA gives it to them in some XML format that is magical, and then Google actually translates it across to GTFS themselves. So, I've been hunting for this data set for like four years, five years now, like since I've been back, and it is correct that it's actually Google's intellectual property because they did the conversion. So, from the data set, from what is released from LTA, they have the schedules for buses, but not for trains. So, for trains I actually had to scrape the various websites, I mean, consolidate manually the various websites. Yeah, and then all that is part, yeah. So, that is, I mean, it would be interesting to actually, the more accurate way of dealing with this is actually reverse engineer the path arrival and train arrival APIs, and then use that to can calculate the travel times. Like, but you're working backwards for that one. Yeah. So, yeah. Question, sir. Is there a way to visualize that in such a way that it's intuitively, okay, so in this one, I just start at the top and you go down on the mountain side. Yeah, yeah. But, you know, intuitively the further you are, the harder it is to get there. Yeah. It seems like there's some kind of inversion that would need to be done. I think the problem is with the analogy. Maybe it is my problem with my analogy if I use the analogy of equating it to topological maps. So, my question is, if you really leave down the top of the mountain, it would be super easy to go down. You want to do the afters. I think it's a problem with the analogy, not particularly with the representation, because I've thought of this particular representation for, I mean, okay, I take it back. There is, there are the guys over at MIT sensible cities that have done it before. What they did is that instead of drawing it in contours, they deformed the space. They respect time, they deformed space. Problemary part about that is that when you do, they did it with driving data, driving and bike data, which is somewhat continuous in terms of the contours. But if you do with public transport, you have islands and it's kind of harder to map that back into this kind of distortion. And it's also not that particularly easy. My point on this is this one is actually more for planning, you've got more for, I wanted to use it more for analysis, rather than for, I find that that one's a lot prettier. Yeah. How do you draw these contours by estimating the vast tree in the weather pack? Okay, so, okay, I lost it. So, what you actually do is that for every, you cut Singapore up into many grids. You cut Singapore up to a, what I did is a 200 by 200 meter grid. Every point pings every other point. So every point you get to every other point in Singapore and you get a travel time from that. So you get kind of this grid and heat map of Singapore, point to many points kind of travel time matrix. And then from the heat map, you essentially draw, you essentially draw, I mean, 12.js and D3S2, you do marching squares. Essentially you do marching squares and you interpolate between points. So that's how you get the contours. Yeah. So, doesn't make sense, doesn't make sense, yeah. Let's see if I have one more question. What's the thing that actually generates the map? Okay, that's a, which part of the map? Okay, the thing that actually generates the, okay, so this is my pipeline. So from, from digging and scraping from GTF, digging and scraping GTFS, I pipe it into OpenTrip Planner, which does the point to point analysis. And then I take the data out and I have an offline script to draw the contours. Cause what comes out of OpenTrip Planner essentially is just a snack, yeah. And then if I really want the contours, and the reason why I need the contours or why I want the contours is because when they live as proper polygons, you can then perform geospatial analysis on it. Like, what's this? Now, you know what actually has this feature? It's just really far, it's hidden really far inside their interface. You can actually pick a couple of places, set a travel time, and then it will filter for the intersection spatially. So that's why I had to move one step further to actually draw the contours cause I wanted the polygons. Yeah. Thanks. Thank you. Thank you. Thank you.