search

GeoDS / LSTM-TrajGAN

LSTM-TrajGAN: A Deep Learning Approach to Trajectory Privacy Protection
68 stars 26 forks source link

How did that syn_best come about #5

Open 18639001046 opened 2 months ago

18639001046 commented 2 months ago

How did that syn_best come about。

suremangood commented 1 month ago

Hello, I would like to ask how your final_test.npy was generated? Is it generated by the csv2npy.py code in the path 'data/' in the project?

sophscha commented 1 month ago

Hello, I would like to ask how your final_test.npy was generated? Is it generated by the csv2npy.py code in the path 'data/' in the project?

@suremangood I found out that the encoding of the test data is slightly different (has 2 features more). Solved it by copying csv2npy.py and adjusting it a little bit for test data:

In line 10 add features 'tid' and 'count' to the list: x = [[] for i in ['lat_lon', 'day', 'hour', 'category', 'mask', 'tid', 'count']]

After line 23 add

x[5].append(tid)
x[6].append(traj.tid.count())
18639001046 commented 1 month ago

谢谢

---- Replied Message ---- | From | @.> | | Date | 10/27/2024 00:11 | | To | @.> | | Cc | @.>@.> | | Subject | Re: [GeoDS/LSTM-TrajGAN] How did that syn_best come about (Issue #5) |

I found out that the encoding of the test data is slightly different (has 2 features more). Solved it by copying csv2npy.py and adjusting it a little bit for test data:

In line 10 add features 'tid' and 'count' to the list: x = [[] for i in ['lat_lon', 'day', 'hour', 'category', 'mask', 'tid', 'count']]

After line 23 add

    x[5].append(tid)
    x[6].append(traj.tid.count())

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>