Travel Mapping
Highway Data Discussion => Updates to Highway Data => Solved Highway data updates => Topic started by: Highway63 on September 07, 2018, 01:22:04 am
-
I could volunteer to do Iowa for you and see where things go from there...
Edit: I went ahead and did Iowa anyway, on my local machine, just to see how long it would take. ;D
18:59, of which 3:48 was just downloading the HighwayData repo, a step that won't need to be repeated.
So, 15:11 for 22 pairs? Not bad. I did leave out many points at state lines, but still wasted time looking them over to see if they could be best fixed by only changing the coords in the IA files. (At that rate, MO would take me... 3:16:42?)
If you'd like, I can put in a pull request and email you the changed files.
YES PLEASE. I spent 45 minutes Tuesday night in the Bloomington area alone and another 20 today on US 150/I-74. I'll hold off trying to plug away at more while you try your method.
Even if that leaves all the state-line points - of which there appear to be quite a few, not all of which are both states I do - that would get me light-years ahead as opposed to my golf-cart speed.
-
YES PLEASE. ...
Even if that leaves all the state-line points - of which there appear to be quite a few, not all of which are both states I do - that would get me light-years ahead as opposed to my golf-cart speed.
Most of the state-line points are between two of your states, which makes things easier. What I'm thinking is:
1st pass: Go state-by-state, only editing NMPs fully within a single state.
2nd pass: Fix NMPs at state lines.
2½: Any NMPs at the MN state line where MN has the obviously better coords.
Whatever's left over could be sorted out with froggie.
I spent 45 minutes Tuesday night in the Bloomington area alone and another 20 today on US 150/I-74. I'll hold off trying to plug away at more while you try your method.
Looks like, since I posted, there have been no changes in IA (https://github.com/TravelMapping/HighwayData/commits/master/hwy_data/IA).
Shall I go ahead and commit my IA changes, then?
Did you have any work in progress there I'd need to watch out for?
Did some work in WI too; I see there have been some (https://github.com/TravelMapping/HighwayData/commits/master/hwy_data/WI/) changes (https://github.com/TravelMapping/HighwayData/commit/414c86eac220b3f78222b159622e0394db4fb1f4#diff-4925eea25faa874ec8a52956cd03fada) there. Will need to give a closer look to make sure none of my edits conflict.
-
^ If Jason's fine with that, I'll take care of the MN/IA border points.
(and now that I look at the log, the MN/WI border points as well)
-
Highway63, are these changes OK with you?
https://github.com/TravelMapping/HighwayData/pull/2201
-
Highway63, are these changes OK with you?
https://github.com/TravelMapping/HighwayData/pull/2201
Those all look good. For the sake of efficiency, since there are so many routes and points and since you have a much better handle on your matching system than I do, it's OK for you to go through the points and submit batches at a time. I just need to know which routes were changed and I'll just download the new versions after the fact (since, as I learned with the Bloomington work, multiple points along a single route can change).
The only weakness I see in this method, and it's hardly a justification for doing it the slow way, is seeing cases when both points are just "off" enough that I want to make a new one for all of them.
-
For the sake of efficiency, since there are so many routes and points and since you have a much better handle on your matching system than I do, it's OK for you to go through the points and submit batches at a time. I just need to know which routes were changed and I'll just download the new versions after the fact (since, as I learned with the Bloomington work, multiple points along a single route can change).
I'm keeping track of which routes change as I go; everything I've saved locally after a certain date/time. If it makes it easier for you, rather than downloading them all one-by-one when it's all said & done, I can ZIP up all the files that have changed and email that to you.
In the meantime, for each pull request, you can click on Files changed (https://github.com/TravelMapping/HighwayData/pull/2201/files) and see which ones.
( ia.i380.wpt ia.ia027.wpt ia.ia141.wpt ia.ia330.wpt ia.us006.wpt ia.us061.wpt ia.us065.wpt ia.us136.wpt ia.us075bussio.wpt il.us030.wpt )
The only weakness I see in this method, and it's hardly a justification for doing it the slow way, is seeing cases when both points are just "off" enough that I want to make a new one for all of them.
When there are only two routes involved in one spot, I usually just take the better-looking coords and copy them into the other file. When there are, say, 3 routes where 2 match and 1 doesn't, I'll keep it simple & edit the 1 file to match the 2 others, unless the 2 routes have coords that are significantly "off" enough for me to decide they're more in need of a fix. If all the points look bad enough, sometimes I'll grab fresh coords from WPTeit and edit all the affected files.
-
Did some work in WI too; I see there have been some (https://github.com/TravelMapping/HighwayData/commits/master/hwy_data/WI/) changes (https://github.com/TravelMapping/HighwayData/commit/414c86eac220b3f78222b159622e0394db4fb1f4#diff-4925eea25faa874ec8a52956cd03fada) there. Will need to give a closer look to make sure none of my edits conflict.
Upon closer look, I see the only change was to WI WI92 WI78, identical to a change I made. We should be in good shape here.
Edit: https://github.com/TravelMapping/HighwayData/pull/2203
-
Didn't see a response from Jason or Eric so assuming I have green light to address the IA/MN and MN/WI border points.
-
How about we wait till the rest of the NMPs in Jeff's states are fixed, to avoid any potential edit conflicts, and then pick off what state line points are left over from there, choosing the best coords for both states as appropriate?
-
Sure, I can wait. I just hadn't seen a response to my query the other day, despite the proliferation of posts.
-
SD NMPs:
https://github.com/TravelMapping/HighwayData/pull/2205
-
I'm keeping track of which routes change as I go; everything I've saved locally after a certain date/time. If it makes it easier for you, rather than downloading them all one-by-one when it's all said & done, I can ZIP up all the files that have changed and email that to you.
That would be great. :) I notice the GitHub file lists don't have a "last modified" option; that's kind of a bummer on general purpose.
When there are only two routes involved in one spot, I usually just take the better-looking coords and copy them into the other file. When there are, say, 3 routes where 2 match and 1 doesn't, I'll keep it simple & edit the 1 file to match the 2 others, unless the 2 routes have coords that are significantly "off" enough for me to decide they're more in need of a fix. If all the points look bad enough, sometimes I'll grab fresh coords from WPTeit and edit all the affected files.
I'd likely do about the same. The whole NMP issue kind of stumps me, though. A handful of the ones I did touch because of this hadn't been touched since at least 2012, when the conversion from GGM to WPT was made.
-
I notice the GitHub file lists don't have a "last modified" option; that's kind of a bummer on general purpose.
The info is there; you just have to surf around a bit to get to it. From the Files Changed (https://github.com/TravelMapping/HighwayData/pull/2201/files) link, click the "View" button in the gray area at the upper-right of the diff for whatever file you want to check out. That lets you view the whole file at the most recent commit. In the case of IA I-380, that takes you here (https://github.com/yakra/HighwayData/blob/d61c2f94ac64a0f7dece165fbaa50825602df11a/hwy_data/IA/usai/ia.i380.wpt). From there, at the right of the gray bar above the .wpt code, click "History". You can see the previous commits (https://github.com/yakra/HighwayData/commits/d61c2f94ac64a0f7dece165fbaa50825602df11a/hwy_data/IA/usai/ia.i380.wpt) for that file, and there are links to the diffs from those commits as well.
The whole NMP issue kind of stumps me, though. A handful of the ones I did touch because of this hadn't been touched since at least 2012, when the conversion from GGM to WPT was made.
As I go thru IL, I'm seeing a lot of Business Interstates. We know Tim wasn't bothered by non-matching points when drafting that system, so there's that. :)
-
Ooh la la! :D
IL NMPs @ https://github.com/TravelMapping/HighwayData/pull/2211
-
MO NMPs @ https://github.com/TravelMapping/HighwayData/pull/2214
Next step = state lines, marking FPs, and anything I may have missed.
-
https://github.com/TravelMapping/HighwayData/pull/2217
froggie, this leaves only
mn.mn271@MN/SD 44.572586 -96.451807
sd.sd028@SD/MN 44.572662 -96.451807
mn.mn019@SD/MN 44.44144 -96.451957
sd.sd030@SD/MN 44.441456 -96.451936
mn.mn060@MN/WI 44.386423 -92.032299
wi.wi025@MN/WI 44.386477 -92.032299
mn.mn269@SD/MN 43.849531 -96.452998
sd.sd011@SD/MN 43.849585 -96.453223
ia.ia139@IA/MN 43.50069 -92.009339
mn.mn139@IA/MN 43.500698 -92.009307
Let me know what you choose to do with these.
-
Pretty much finished editing files. PM me your email and I can zip them up and send.
-
There are new NMPs:
ia.ia139@IA/MN 43.50069 -92.009339
mn.mn139@IA/MN 43.500698 -92.009307
mn.mn060@MN/WI 44.386423 -92.032299
wi.wi025@MN/WI 44.386477 -92.032299
mn.mn019@SD/MN 44.44144 -96.451957
sd.sd030@SD/MN 44.441456 -96.451936
mn.mn271@MN/SD 44.572586 -96.451807
sd.sd028@SD/MN 44.572662 -96.451807
mn.mn269@SD/MN 43.849531 -96.452998
sd.sd011@SD/MN 43.849585 -96.453223
And in MO: http://forum.travelmapping.net/index.php?topic=2377.msg14198#msg14198
-
Do these still need to be addressed? I appear to have glossed over this part.
-
Do these still need to be addressed? I appear to have glossed over this part.
There are more now:
http://travelmapping.net/logs/nmpbyregion/ia.nmp
ia.ia139@IA/MN 43.50069 -92.009339
mn.mn139@IA/MN 43.500698 -92.009307
ia.i029@143 42.434441 -96.374967
ia.us075bussio@I-29(143) 42.434322 -96.374774
ia.us006@1stAve_Cor 41.669049 -91.565788
ia.us006hisiow@1stAve_S 41.669129 -91.565895
ia.i029@49 41.232154 -95.851443
ia.i080@*49(29) 41.232114 -95.8514
ia.i029@49 41.232154 -95.851443
ia.us006@*I-29(49) 41.232114 -95.8514
ia.i029@50 41.232413 -95.879273
ia.i080@*50(29) 41.232413 -95.87923
ia.i029@50 41.232413 -95.879273
ia.us006@*I-29(50) 41.232413 -95.87923
http://travelmapping.net/logs/nmpbyregion/wi.nmp
mn.mn060@MN/WI 44.386423 -92.032299
wi.wi025@MN/WI 44.386477 -92.032299
http://travelmapping.net/logs/nmpbyregion/sd.nmp
mn.mn019@SD/MN 44.44144 -96.451957
sd.sd030@SD/MN 44.441456 -96.451936
mn.mn271@MN/SD 44.572586 -96.451807
sd.sd028@SD/MN 44.572662 -96.451807
mn.mn269@SD/MN 43.849531 -96.452998
sd.sd011@SD/MN 43.849585 -96.453223
http://travelmapping.net/logs/nmpbyregion/il.nmp
il.il004@CR40 39.676285 -89.704528
il.us066hiscar@IL4Cha_N 39.67631 -89.704485
il.il004@WriSt 39.485197 -89.776368
il.us066hiscar@SneRd 39.484965 -89.776583
il.il003@IL150_E 37.914138 -89.820786
il.il150@HolSt 37.914087 -89.820872
http://travelmapping.net/logs/nmpbyregion/mo.nmp
mo.i270@26B 38.775499 -90.334654
mo.us066hisflo@I-270(26B) 38.775934 -90.334911
mo.i270@32 38.769778 -90.220413
mo.us066hisflo@I-270(32) 38.770246 -90.22067
mo.mo163@*US63 38.868733 -92.252047
mo.us063@*OldMO163 38.868733 -92.251682
Just let me know if any pair is false positive.
-
The Iowa points regarding I-80, I-29, and US 6 with I-29/80 (the bottom 8 on the list) are FP's - I intentionally broke those because of the move of I-29 and I-80 to the dual divided freeway.
The starred points for MO 163/US 63 are FP's; the intersection is closed and MO 163 now turns south, so I wanted to get it close.
The MO US 66/I-270 points are FP's; 66 follows a frontage road.
All the MN border-related points need fixed.
First pair of IL 4/US 66 needs fixed.
Second pair of IL 4/US 66 is FP - but I have a duplicate label issue on 66 and the SneRd after point IL4_AubD needs to be renamed CurRd.
The IL3/150 pair I need to take care of because I didn't finish the job of relocating IL 3.
-
Fix the errors. When only FPs are left (the files are updated with every site update), let me know, I can mark them FP if you are not sure about the procedure (it's not as simple as for "normal" data check errors).
-
Moving from other topic:
The MO 66 Hist/I-270 issue is a false-positive mismatch: 66 is on Dunn Road, close enough to be the frontage road. I named the points after the interchanges because I thought that would make it more convenient. If you disagree, I can rename 26B->GraRd and 32->BelRd.
US 63/163 also false positive.
Again, let me know when all are fixed and I'll mark the remaining NMPs as FP.
-
I've sent all the relevant new files. If something still looks askew after that let me know.
-
Iowa: http://travelmapping.net/logs/nmpbyregion/ia.nmp
The Iowa points regarding I-80, I-29, and US 6 with I-29/80 (the bottom 8 on the list) are FP's - I intentionally broke those because of the move of I-29 and I-80 to the dual divided freeway.
I've marked the broken concurrency of I29 to I80/US6 in Omaha FP, should be in with the next site update: https://github.com/TravelMapping/HighwayData/pull/3490
The NMP at I29 wp 143 with US75 Bus Sio needs to be fixed. I recommend changing the US75Bus Sio wp.
The NMP b/n US6 and US6 His Iow in Iowa City needs to be fixed. I recommend changing the US6 His Iow wp. A to do for @Si? not sure who is doing what here....
Wisconsin: http://travelmapping.net/logs/nmpbyregion/wi.nmp
Currently fine!
South Dakota: http://travelmapping.net/logs/nmpbyregion/sd.nmp
Currently fine!
Illinois: http://travelmapping.net/logs/nmpbyregion/il.nmp
Still 2 NMPs at IL4 and US66 His Car:
il.il004@CR40 39.67631 -89.704573
il.us066hiscar@IL4Cha_N 39.676285 -89.704528
il.il004@WriSt 39.485197 -89.776368
il.us066hiscar@SneRd 39.484965 -89.776583
First pair of IL 4/US 66 needs fixed.
The first pair just needs to be lined up. I'd change the US66 point. A to do for @Si? not sure who is doing what here....
Second pair of IL 4/US 66 is FP - but I have a duplicate label issue on 66 and the SneRd after point IL4_AubD needs to be renamed CurRd.
IL4's WriSt is at the railway but not at Writhman Street where it should be.
US66's SneRd is off.
IL4 misses SneRd wp at all.
US66 misses WriSt wp at all.
Missouri: http://travelmapping.net/logs/nmpbyregion/mo.nmp
The starred points for MO 163/US 63 are FP's; the intersection is closed and MO 163 now turns south, so I wanted to get it close.
The MO US 66/I-270 points are FP's; 66 follows a frontage road.
Marked FP: https://github.com/TravelMapping/HighwayData/pull/3490
-
Not NMPs but.... I think that the HIDDEN_JUNCTION (http://travelmapping.net/devel/datacheck.php?rg=IA,IL,MO,SD,WI) errors are just because IL3 misses IL46's CavRd wp and IL111's ToUS67 wp.
US66 His FR144 wp is just not a visible wp on MO266 but coordinates are lined up.
-
Illinois: http://travelmapping.net/logs/nmpbyregion/il.nmp
Still 2 NMPs at IL4 and US66 His Car:
il.il004@CR40 39.67631 -89.704573
il.us066hiscar@IL4Cha_N 39.676285 -89.704528
il.il004@WriSt 39.485197 -89.776368
il.us066hiscar@SneRd 39.484965 -89.776583
First pair of IL 4/US 66 needs fixed.
The first pair just needs to be lined up. I'd change the US66 point. A to do for @Si? not sure who is doing what here....
Second pair of IL 4/US 66 is FP - but I have a duplicate label issue on 66 and the SneRd after point IL4_AubD needs to be renamed CurRd.
IL4's WriSt is at the railway but not at Writhman Street where it should be.
US66's SneRd is off.
IL4 misses SneRd wp at all.
US66 misses WriSt wp at all.
Consarn it, I thought I had that first one taken care of. (All routes in my states should go through me.)
As for the point(s) south of Virden, I didn't think it made sense to have points immediately before and after the railroad crossing, and think one could do double duty - located at the RRX, splitting the difference. However, in thinking of that, I neglected to make sure that Hist 66 and 4 converged at the same place.
And I'll fix IL3's missing waypoint with IL 146 and the ToUS67 issue.