traffic variable

[temospena]

Considering that googetraffic package only allows to retrieve live information, and in raster, it would require to "wait" for some hours for the results of transit dynamics.

On the other hand, the mapboxapi allows for requests of "typical" day and time (intervals of 15min), but with the cost of the max API requests per month. The results are retrieved in vector.

I would suggest to explore the mapboxapi package, and make the requests at the final stage, for the cell candidates, and not for the whole city limit. This would lower the request amounts, and maybe provide a faster process.

More info:

• https://github.com/dime-worldbank/googletraffic
• https://www.mapbox.com/traffic-data
• https://labs.mapbox.com/what-the-tile/
• https://walker-data.com/mapboxapi/

[temospena]

This is only classifying for one timestamp!

It is considering that if at least 10% of total segment length is with heavy or severe congestion, it is congested.
Does it makes sense, @valenca13

Recalling the paper definition

if a zone has a street with high or heavy traffic levels at any time of day, there is a dispute for space, which increases the complexity of reallocating it. We calculate the maximum traffic level of service of every street within each zone from 6 AM to 11 PM. Only zones where the maximum value of the level of service is high (level 3) or heavy (level 4) are selected as a requirement for complex zones to reallocate road space. This means that we select zones that have at least one street that intersects the zone, with high (level 3) or heavy (level 4) traffic levels at any time of day.

[valenca13]

This is only classifying for one timestamp!

It is considering that if at least 10% of total segment length is with heavy or severe congestion, it is congested. Does it makes sense, @valenca13

Recalling the paper definition

if a zone has a street with high or heavy traffic levels at any time of day, there is a dispute for space, which increases the complexity of reallocating it. We calculate the maximum traffic level of service of every street within each zone from 6 AM to 11 PM. Only zones where the maximum value of the level of service is high (level 3) or heavy (level 4) are selected as a requirement for complex zones to reallocate road space. This means that we select zones that have at least one street that intersects the zone, with high (level 3) or heavy (level 4) traffic levels at any time of day.

I think it does make sense, but this can be discussed. The idea here is to analyze if there is any conflict during the day. @temospena

[temospena]

But how to define "any"?
Looking at the cell on the right, is that very little green segment enough to classify the whole cell as with conflict?

(Imagining that this is the max result of congestion of all segments during all day).

[temospena]

Is there a possibility to use the road hierarchy levels (primary, secondary, tertiary - excluding residential, off road and links), as a proxy of this variable?
Assuming that these kind of high level roads are somehow congestioned once a day everywhere...

Something to think and discuss.
(This would be much simple to incorporate in the process.)

[temospena]

Something like this, considering primary, secondary and tertiary highway tags as possibly congested.