A package used in DNN trainning in ATLAS analysis
Project description
Python package use cuda to normalize input variables using cuda package in ATLAS analysis
Function use to do Guassian Normalization: Mean: $$\mu_{i}=\frac{\sum x_{i}\times w_{i}}{\sum w_{i}}$$ Variance: $$\sigma_{i}=\frac{\sum (x_{i}\mu_{i})^{2}\times w_{i}}{\frac{N1}{N}\times\sum w_{i}}$$ Normalized input feature: $$\bar{x_{i}}=\frac{x_{i}\mu_{i}}{\sigma_{i}}$$
Main function: guass_normal((1),(2),(3))
Input:
(1):Numpy array contain all input features you want to normalize. (2):Numpy array used to calculate each feature's mean and variance. (3):1d Numpy array contains each events weight in (2)
(1) and (2) must have the same number of columns.
cuda_cut((1),(2),(3)): Used to calculate event yield after applying DNN cut.
Input: (1): 1d numpy array include the variable you want to cut. (2): 1d numpy array include event weight. (3): cut threshold
Project details
Release history Release notifications  RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for cuda_guass_normal1.13py3noneany.whl
Algorithm  Hash digest  

SHA256  4bf6b5b1af2a25363fc0de1b3ac4cedede59e0dd840796abd6befde81f425437 

MD5  17e2bb26b882b588814def84b5035023 

BLAKE2256  32bc0f0acb2a2da6a13341fc2db1a1b01b631b4ed68af3f8b0543e188853ff24 