rongxiangsu / mobilitymotifs Goto Github PK

mobilityMotifs

Datasets:

Preprocessed survey data from Chicago to run the survey motif detection algorithm: Chicago

Simulated phone usage to run the phone motif detection algorithm: Example

Source files:

The c++ file for the introduced perturbation model for the Chicago survey: Chicago Input: Models the motifs for a given size (parameter: numbernodes in the source-file) Output: a) motifs_N.dat: fraction of the most common daily networks and the total number of networks with size N if they appear more than 10 times b) degreedistribution.dat: size distribution of the motifs c) edgedistribution.dat: edge distribution of the motifs d) travel.dat: number of trips between different activities e) traffic.dat: trips per 30min-interval f) predictability.dat: fraction of time spend at work and home g) duration_home.dat: distribution of consecutive time spend at home h) duration_work.dat: distribution of consecutive time spend at work i) duration_other.dat: distribution of consecutive time spend at other locations j) intertime_home.dat: distribution of time between two home locations
k) intertime_work.dat: distribution of time between two work locations
l) intertime_other.dat: distribution of time between two other locations
Daily networks are shown at the end of the program (not in a file), if they appear more than 10 times. The structure of the networks is the following: Motif_ID total_count fraction Motif_Matrix (N x N)

The c++ file for motif detection in the survey data: Chicago Input: Calculates the motifs for a given size from data.dat (parameter: numbernodes in the source-file) Output: a) degreedistribution_weekday.dat: size distribution of the motifs b) distances_weekday_N.dat: distribution of the reported distance of the daily path vs the optimal distance of the daily path visiting the same locations for motifs with the given size N c) distances_weekday_com_N.dat: cumulative distribution of distances_weekday_N.dat d) motifs_weekday_N.dat: motif ID and the fraction of these motifs and the total number of motifs with size N

Daily networks are shown at the end of the program (not in a file), if they appear more than once. The structure of the networks is the following: Motif_ID total_count fraction Motif_Matrix (N x N)

The c++ file for motif detection in the simulated phone usage: Example Input: Calculates the motifs for a given size from data_example.dat Output: a) unknowndis.dat: distribution of unknown 30-minutes intervals for daily networks b) degreedis_weekend_min8e.dat: size distribution of the motifs for weekends c) degreedis_weekday_min8e.dat: size distribution of the motifs for weekdays d) motifs_min8e.dat: motif distribution for any day e) motifs_weekend_min8e.dat: motif distribution for weekends f) motifs_weekday_min8e.dat: motif distribution for weekdays g) transition_weekendday_relative_min8e.dat: normalized transition probability between motifs for any days h) transition_weekendend_relative_min8e.dat: normalized transition probability between motifs for weekends i) transition_weekdayday_relative_min8e.dat: normalized transition probability between motifs for weekdays j) transition_weekendday_absolut_min8e.dat: total number of transitions between motifs for any days k) transition_weekendend_absolut_min8e.dat: total number of transitions between motifs for weekends l) transition_weekdayday_absolut_min8e.dat: total number of transitions between motifs for weekdays

Motifs_IDs are from the paper