Waveform-prediction

Harnessing Machine Learning Framework for Data-Driven Predictive Waveform Optimization in Piezoelectric Inkjet Printing Utilizing Multi-Material Inks for Enhanced Droplet Control

Conda Environment Setup Guide

This guide provides instructions for installing Conda, creating a Conda environment, and installing the required dependencies using the provided environment.yml file, and running the deep learning (DL) and macchine learning (ML) models.

Instructions to Install Conda and Set Up the Environment

Step 1. Install Conda:

• Download the installer for Miniconda or Anaconda:
  • Miniconda: https://docs.conda.io/en/latest/miniconda.html
  • Anaconda: https://www.anaconda.com/products/distribution
• Run the installer and follow the setup instructions for your operating system.
• Verify the installation: conda --version

Step 2. Open an Anaconda Prompt terminal.

Step 3. Navigate to the directory where environment_CPU.yml is saved: cd /path/to/directory

Step 4. Create the Conda environment:

• For the CPU environment: conda env create -f environment_CPU.yml
• For the GPU environment: conda env create -f environment_GPU.yml

Step 5. Verify the environment is created: conda env list

Step 6. Activate the environment:

• For the CPU environment: conda activate tensor
• For the GPU environment: conda activate tensorGPU

Step 7. Verify that the required libraries are installed: conda list pip list

Notes:

• Replace /path/to/directory with the actual path where the .yml files are located.
• If you encounter issues with pip packages, ensure pip is installed and updated: conda install pip pip install --upgrade pip
• The GPU libraries are configured for Nvidia GTX 1650. For other GPUs, refer to the TensorFlow compatibility guide to adjust the library versions: https://www.tensorflow.org/install/source
• Experimental dataset used in this article is available online at https://www.eecs.yorku.ca/~grau/MLIJ/Supplementary%20material.zip.

File Descriptions:

Following files can be run using Jupyter-lab under the Anaconda environment as described in above section.

• data_split.ipynb This file helps to concatenate individual dataset files to a single file, and then splitting that file into train and test datasets.
• DL_model_train_test.ipynb The data pre-processing can be performed on the dataset, and then DL model can be trained and evaluated on preprocessed data file using this script.
• GSCV_ML_model_train_test.ipynb The data pre-processing can be performed on the dataset, and then a GridSearch operation be performed on the ML models to find the best optimal hyperparameters of each model.
• ML_model_train_test.ipynb The data pre-processing can be performed on the dataset, and then ML models can be trained and evaluated on preprocessed data file using this script.