ASMAD - See Average Delays by Station for an Entire Route

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Status
Not open for further replies.

acelafan

Conductor
Joined
May 24, 2009
Messages
1,064
Location
Atlanta, GA
As most of you know, ASMAD is the Amtrak Status Maps Archive Database. It lets you search John's status maps files in a variety of ways to see past Amtrak delays/lateness.

I have finally finished a new search feature that I've been developing for awhile now. The results mimic an individual Amtrak Status Map file for a particular train, but it creates a bar chart of the average delay for each station along the route. Here is the main link to the page.

This page requires the least number of clicks to see the data, just enter the train number and a bar chart will appear showing the average arrival and departure delay for each station along the route for the past 30 days. Click an individual bar and you will get the "history chart" showing past delays at just that station for the same time period.

A summary table is provided under the bar chart with the specific delays listed by station. You can change the date range as you wish and also filter by specific days like the other search pages.

Because there will be some missing reports in the Amtrak data, I included a column called "available data". Basically 100% means no missing reports; lesser values indicate some trains did not report their status for a particular station over the date range that you selected. You'll get the hang of it.

Here is an example bar chart showing delays for the westbound Zephyr over the past 30 days since it's been newsworthy:

20150708_average_7_chart.png


And the table:

20150708_average_7_table.png


This is the most complex page I've coded on ASMAD. Every time I went to declare victory I'd find a minor bug but I think they are all worked out now.

Hope you find it useful for planning or general info.

Chris J
 
Very impressive! Although the data displayed, especially for western LD trains, is downright depressing. :angry2:
 
Really neat! I think I'll take a look at our departure station,WPK, based on the data I won't worry so much about cross town traffic getting there.
 
49 shows a pattern: time lost on the Empire Corridor, recovered west of there, lost again in Indiana.
 
I have finally finished a new search feature that I've been developing for awhile now. The results mimic an individual Amtrak Status Map file for a particular train, but it creates a bar chart of the average delay for each station along the route. Here is the main link to the page.
Thanks for the neat new feature. Very useful way to see the pattern of where the delays tend to occur on many routes.
I'm enough of a stats geek that I wonder if the median delay rather than the mean average would look different for some trains, particularly over a shorter, say 1 month period. A couple of trips with 6 hour delays due to a broken down freight train, grade crossing collision, or a severe weather event would pull the mean average delay up while the median would be less.
 
Very impressive! Although the data displayed, especially for western LD trains, is downright depressing. :angry2:
The Coast Starlight is doing OK in recent months for OTP. The SWC, while not good, has been doing better than the CZ, TE, EB with the CZ taking the lead in the contest for worse summer for really late LD train for 2015. In the east and midwest, the NS meltdown has not returned for the CL and LSL, and the CONO has been quite nicely for OTP in recent weeks. So it is not all bad for severe constant delays for the LD trains this year.
 
I have finally finished a new search feature that I've been developing for awhile now. The results mimic an individual Amtrak Status Map file for a particular train, but it creates a bar chart of the average delay for each station along the route. Here is the main link to the page.
Thanks for the neat new feature. Very useful way to see the pattern of where the delays tend to occur on many routes.
I'm enough of a stats geek that I wonder if the median delay rather than the mean average would look different for some trains, particularly over a shorter, say 1 month period. A couple of trips with 6 hour delays due to a broken down freight train, grade crossing collision, or a severe weather event would pull the mean average delay up while the median would be less.
You are right, the median would be quite different especially when trains are extremely late. When I first started the average delays page I thought about trying to include median but ran out of time. Something for the future, perhaps.

I've had to remind myself how to find the median, and a good explanation is at https://www.mathsisfun.com/median.html
 
This new feature is another way to see the extent of Amtrak's poor on-time-performance. It's discouraging to see the Lynchburger for June ran average of almost half an hour late. Most of the delays happen on NS south of Alexandria. It's the same story for the other train (89) I've taken lately = the delays begin south of WAS. I think VA has invested a lot of money for this to still be happening to these trains.
 
This new feature is another way to see the extent of Amtrak's poor on-time-performance. It's discouraging to see the Lynchburger for June ran average of almost half an hour late. Most of the delays happen on NS south of Alexandria. It's the same story for the other train (89) I've taken lately = the delays begin south of WAS. I think VA has invested a lot of money for this to still be happening to these trains.
However, if you look at the on-time results for Regionals 145-171 (to get all 3 trains) arriving at LYH for the month of June, there was a block of circa one hour plus late arrivals from June 7 to June 15 which raised the average. Spot checking the late days, the delays ere mostly between ALX and CVS. The only on-time day in that block was on Sunday. Which suggests track work might have been the cause. Interesting test case for a median delay average as it would have been lower.

The average delay for Regional 145-171 arriving at LYH since June 16 is 13 minutes which is decent.

I don't believe VA contributed much in track improvements prior to starting the Lynchburger service, mostly for a layover track and improvements in LYH. Virginia is investing funds for improvements to the NS tracks from ALX to LYH. VA is contributing $6.4 million over FY15, FY16 to a $9.2 million project titled "Lynchburg to Alexandria Speed Improvements". What the speed improvements are, don't recall seeing any details on it. VA is also contributing $22.1 million in FY15, FY16 to a $31.6 million project to double track a single track segment from Nokesville to Calverton (south of Manassas). Don't know when the track work will actually get done, but I would venture that VA DRPT might be seeking to complete these projects prior to or shortly after the start of the service extension to Roanoke for faster LYH and Roanoke trips.
 
I have finally finished a new search feature that I've been developing for awhile now. The results mimic an individual Amtrak Status Map file for a particular train, but it creates a bar chart of the average delay for each station along the route. Here is the main link to the page.
Thanks for the neat new feature. Very useful way to see the pattern of where the delays tend to occur on many routes.
I'm enough of a stats geek that I wonder if the median delay rather than the mean average would look different for some trains, particularly over a shorter, say 1 month period. A couple of trips with 6 hour delays due to a broken down freight train, grade crossing collision, or a severe weather event would pull the mean average delay up while the median would be less.
This wasn't too hard, I finished at lunch but didn't get a chance to post the change til now.

Calculating the median delay is now an option on the average delays page. The value is also included on the other pages where I had plain average included at the bottom of the table.

As you thought, it is lower in cases of extreme delays.
 
I love the new feature. Functionality is great, data not so much (although it's accurate) - thanks!
 
This wasn't too hard, I finished at lunch but didn't get a chance to post the change til now.

Calculating the median delay is now an option on the average delays page. The value is also included on the other pages where I had plain average included at the bottom of the table.

As you thought, it is lower in cases of extreme delays.
Thanks for adding the Median delay option. I did comparisons for some LD trains over the past month. Not much difference between the mean and median for LSL #49 which reflects how it routinely runs late across upper NY. However, the Cardinal #50 shows a noticeable difference in the pattern between the mean and median. The Median delays are much lower from CHI to WV, so a good portion of the #50 runs are close to on-time on that leg. Then both the median and mean delays increase going through WV and on the Buckingham Branch railroad. However, as the VA funded maintenance and improvement projects to the BB are completed, the Cardinal's time-keeping on the BB have improved over the past several years. although it still often runs late on the BBRR.

Another LD train that is interesting is the Coast Starlight #11. Not for the difference between the mean and median delay, but that its accumulated delays over the past month have mostly been from SEA to PDX and the overall delay trend is downward south of PDX. So when the Cascades corridor work is done in 2017, the southbound Coast Starlight OTP should improve overall.
 
Status
Not open for further replies.
Back
Top