search

ComputationalMovementAnalysis / utterances

A place to hold the comments made via utterances
1 stars 0 forks source link

W03_04_tasks_and_inputs #2

Open utterances-bot opened 3 years ago

utterances-bot commented 3 years ago

Tasks and Inputs | Computational Movement Analysis: Patterns and Trends in Environmental Data

Tasks and Inputs | Computational Movement Analysis: Patterns and Trends in Environmental Data

https://computationalmovementanalysis.github.io/W03_04_tasks_and_inputs.html

mariottomarina commented 3 years ago

Question: Input for Similarity measures in Task 6

I'm stuck at Task 6. Maybe some of you can help me. So far I was able to seperate the pedestrian file into seperate matrices (one matrix for each trajID). But I still have the columns TrajID, E, N and DatetimeUTC. When I try to compute the similarity measures with the matrices, I get an error message. I think this input for the functions is wrong as there can only be one vector for traj1 and one for traj2 (i guess). My problem is that i don't know which values i need to keep and which to toss for the calculations to work. Does anyone know which values I need to feed the function?

nic-steinm commented 3 years ago

Hi Marina, I used the select() function to select the columns E and N. I hope this helps :)

nic-steinm commented 3 years ago

Oh, and then you still need to convert your data table into a matrix so the algorithms can read them.

ratnanil commented 3 years ago

As @nic-steinm mentioned, the functions cannot handle Datetime (and therefore assumes that both trajectories are sampled concurrently) and are passed as a matrix.

ratnanil commented 3 years ago

An alternative would be to convert DateTimeUTC into integer (POSCXct is nothing else than the number of seconds after a specific date). All functions allow multidimensional inputs, so the third dimension would be time