How is it possible to generate the datasets for multiple runs/trials?

[lenardxu]

I'd like to use your function of running trajectory evaluation for multiple trials given the specific algorithm and dataset, by enabling the argument --mul_trials. However, what I want to ask is how did you generate the estimate files of several trials (e.g., "stamped_traj_estimate0.txt", "stamped_traj_estimate1.txt", "stamped_traj_estimate2.txt", ...) corresponding to the same ground-truth file? After checking your template trajectory files, I still have no clue of generating them in experiment, but only able to generate one trial as usual. Or is it only possible to generate such datasets in simulation?
Many thanks in advance!

[deepak046]

I have the same issue as you. Due to this I'm not able to generate plots for euroc_mono_stereo dataset as well as my own datasets since I just have a single trajectory estimate .txt file

[deepak046]

Nvm. I figured it out. The --mul-trials is meant to be used if we have data from different trials runs of our algorithms. So just remove the flag and it defaults to None and it then works.