image
imagewidth (px) 640
640
|
---|
This is a dataset for two-tx simulations
We have two h5 files:
-train.h5
-test.h5
train has 100000 and test has 20000 randomly generated two-tx data. This data is stored in string keys (there was no other way :( ) corresponding to their index. Example : "10", "1001", "0"
H5 row contains all hitpoints in n rows and 5 columns where n is the number of hits. These columns are : "x,y,z,hit_time(in seconds),hit index (which tx particle came from)" Also every row contains "config" attribute where it stores necessary info like tx_centers of molecule_counts
I also provided raw one-tx experiments as json file. They are stored with integer indexes. After loading with data = json.load(filename)
you can get hit data and config by:
data[index]["hit_data"],data[index]["config"]
I will provide code to read h5 data easily
I am going to update upper part.
hard_train have 40000 experiments with 0 <= angle < 30 20000 experiments with 30 <= angle < 45 20000 experiments with 45 <= angle < 90 20000 experiments with 90 <= angle < 180
Same with hard_test
- Downloads last month
- 42