SPRINT-imputations 0.0.2

Using SPRINT we do the the impuatation

SPRINT

Using SPRINT we do the the imputation. So, when a dataset having absenceny (or NaN), is processed by this algorithm we can get a complete dataset.

Installation

You can install the package using below command

pip install 

Usage

First import the package

import SPRINT_imputation

Call the function 'SPRINT'

SPRINT_imputation.impute()

Agruments for impute():

• Mandatory
  • Original_Dataframe
  • Tens_shape
• Optional
  • zero_as_missing
  • miss_type
  • missing_rate
  • missing_rate_val
  • block_window
  • ranks
  • max_iter (Keep above 100)
  • hy1_list
  • hyp_prior_list
  • hyper_smooth
  • Batch_per
  • Batch_iter
  • p

Output

A matrix (or dataset), without containing any NaNs.

:bulb:TIP: Regarding CUPY library

• Do note that the CUPY library is not mentioned in the 'requirements.txt' and hence is needed to be installed sperately whose detials steps are mentioned here.

• The CUPY library is suppose to be install as per the GPU's driver version more info can be found here

• If there is any CUDA PATH issue, one can explictly set the PATH by the following command.

    
    import os 
    os.environ['CUDA_PATH'] = "C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v11.7"

:heavy_exclamation_mark: Warning

Don't forgot to change the CUDA version in the above code as per your system.

Contact

Shubham Sharma | Higher Degree Research Scholar (Ph.D.)
School of Civil and Environmental Engineering
Faculty of Engineering | Queensland University of Technology
Mail ID : [e-mail]