udayrage / partialperiodicspatialpatterns

partialPeriodicSpatialPatterns

Partial Periodic Spatial Pattern Mining (PPSPM) aims to find all neighboring itemsets that occur at regular intervals in a spatiotemporal database.

The algorithms, n_ECLAT and ST_ECLAT, have been written in Python 3. The command to execute the program is as follows.

python3 ST_ECLAT.py maxIAT minPS temporalDatabaseFile item_NeighborsFile outputFileName

python3 n_ECLAT.py maxIAT minPS temporalDatabaseFile item_NeighborsFile outputFileName

The maxIAT and minPS values are to be specified in count.

To test the repetability of our experiments, we have provided some databases in the Datasets folder. The details are as follows:

1. T10I4D100K.txt is a synthetic database used in our experiments. The first column represents the timestamp (or transactional identifier) and remaining columns represents items.
2. coordinates.txt file contains the coordinates of the items in T10I4D100K.txt file. Please note that the line number implicitly represents the item number in T10I4D100K.txt. E.g. the coordinates (x,y) in the first line represent the spatial coordinates for the item whose id is 1 in datat10.txt file.
3. dist_Threshold.txt contains the information regarding the items and their neighbors. The first column in each row represents an item i and remaining columns represents the neighbors of i whose distance is no more than the user-specified threshold value.
4. neighbourGenerator.py is another python file used to create dist_Threshold.txt from coordinates.txt file. For brevity, Euclidean distance is used to compute the distance between the items. The command to execute this file is as follows

   python3 neighbourGenerator.py inputCoordinatesFile maxDist outputFileName

Examples of running above code

python3 ST_ECLAT.py 6000 50 Datasets/T10I4D100K.txt Datasets/dist_5.txt patterns_dist_5.txt

python3 n_ECLAT.py 6000 50 Datasets/T10I4D100K.txt Datasets/dist_5.txt patterns_dist_5.txt

/usr/bin/time -v python3 ST_ECLAT.py 6000 50 Datasets/T10I4D100K.txt Datasets/dist_5.txt patterns_dist_5.txt (ubuntu)

/opt/local/libexec/gnubin/time -v python3 ST_ECLAT.py 6000 50 Datasets/T10I4D100K.txt Datasets/dist_5.txt patterns_dist_5.txt (Mac)