Possible "Data Blackout" in the Matrix API

Hi guys.

I’m using ORS a lot recently for a research in a master’s degree and often come back to this forum where I’ve been learning a lot, so I thank you all in advance.

So, I’m using Matrix V2 to obtain a lot of distances and apparently in all consultations that I’ve performed seems to be a “Data Blackout” in the retrieved Matrix, when a lot of empty values and zeros appear in a sequence and then the data returns to “normal values” again. These blackouts occur multiples time in a single consultation.
Here is a example of what is happening…

So, do you guys have any ideia of how can I solve this “error”?! Or how can I treat this?!

Best Regards

Hi,

thanks for your feedback. Could you maybe provide the following additional information:

  1. Is the described phenomenon reproducible across your requests, i.e. does a given requests always fail in the same way?

  2. What tool are you using for querying our API, and how do you handle the response?

  3. Give a specific example of a request that fails for you?

Cheers,
Andrzej

Hello.

About your questions:

  1. Yes, all of my requests, independent of the size of the matrix, presentes these blackouts.

  2. I’m using an algorithm that I wrote in python and using coordinates obtained from Google or Bing, since the limit os the ORS geocoding is too low for my needs. But in both cases the errors remain.


3. The previous table has thousands of coordinates and I use the matrix to obtain the distances between the first eight coordinates (collumns in response table) to the remaining coordinates (rows in response table). Follows another example of the “blackout” (fist 50 rows) in the response table. These blackouts occur dozens of times along the table.

If it helps I can send you the many tables with data that I retrieved from ORS and even my algorithm. Anyway, I am able to help as well as I can to improve the system if it is in fact an error.

Best regards.
Thiago.

Please, disconsider the last image that I submited (screenshot 3) because the are the data that I treat to my research: distances bigger than 50000m are put to zero.

So, the original response table with the “blackout” in the fist 50 rows is:

Hi guys.

Good news: I think that I just found out what was happening.
I checked the data and I noted that these blackouts only ocurred in consultations blocks (due to the ORS limit of 50 locations per consultation) that had coordinates too diferent for the rest of the data. Example: I am in the south hemisphere (negative lat) and there was coordinates with positive lat, - north hemisphere.
So, it seems that due to the restriction of the distance as in https://openrouteservice.org/restrictions/ instead of the response data appears with error only in the respective coordinate, all the data remaining in the consultation block with an coordinate too diferent comes with erros.

To solve it I just applied a filter excluding coordinates too diferent from the majority of the coordinates, and these blackouts never happened again. = ]

I once again thank you all for the support and disponibility to help. I hope that my doubt could help others with a similar issue.

Best Regards.
Thiago