We provide codes and sample data for ResAct.

Codes:
1. raw_data_process.py contains functions to process data in hive table.
2. data_to_buffer.py contains funtions to packaging processed to replay buffer for training.
3. Implementations of ResAct and the baselines are in ./algos folder, and the corresponding configuration files are in ./config
4. run example: python main.py --config=resact with lr=0.00005 probe_num=20

Data:
Since the whole dataset is very large, we cannot upload it the system. Thus we provide samples of 100 user data in the file ./data_samples/user_data