Metadata-Version: 2.1
Name: easycheml
Version: 0.2.1
Summary: A simple tool for using artificial intelligence in chemistry
Home-page: https://github.com/anandsahuofficial/easycheml.git
Author: Anand Sahu
Author-email: anandsahuofficial@gmail.com
License: UNKNOWN
Download-URL: https://github.com/anandsahuofficial/easycheml/archive/refs/tags/easycheml.tar.gz
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Description-Content-Type: text/markdown

ReadMe
======

.. .. image:: https://badge.fury.io/py/easycheml.svg
..         :target: https://badge.fury.io/py/easycheml
..         :alt: PyPI version
.. .. image:: https://travis-ci.com/pycroscopy/easycheml.svg?branch=master
..         :target: https://travis-ci.com/pycroscopy/easycheml
..         :alt: Build Status
.. .. image:: https://readthedocs.org/projects/easycheml/badge/?version=latest
..         :target: https://easycheml.readthedocs.io/en/latest/?badge=latest
..         :alt: Documentation Status
..         :alt: Codacy Badge
.. .. image:: https://pepy.tech/badge/easycheml/month
..         :target: https://pepy.tech/project/easycheml/month
..         :alt: Downloads


.. .. image:: https://colab.research.google.com/assets/colab-badge.svg
..         :target: https://colab.research.google.com/github/pycroscopy/easycheml/blob/master/examples/notebooks/Quickstart_easycheml_in_the_Cloud.ipynb
..         :alt: Colab
.. .. image:: https://img.shields.io/badge/Gitpod-ready--to--code-blue?logo=gitpod
..         :target: https://gitpod.io/#https://github.com/pycroscopy/easycheml
..         :alt: Gitpod ready-to-code

What is EasyCheml
--------------
EasyCheml is a tensorflow-based package for deep/machine learning application in chemistry, which doesn't require any advanced knowledge of Python (or machine learning).  The intended audience is domain scientists with basic knowledge of python.

Why EasyCheml
^^^^^^^^^^
The purpose of the EasyCheml is to provide an environment that bridges the instrument specific libraries and general physical analysis by enabling the seamless deployment of deep/machine learning algorithms.

How to use it
-------------

Data Preprocessing
^^^^^^^^^^^^^^^^^^^^^^

>>> from easycheml.preprocessing import PreProcessing as p
>>> preprocessed_dataset,train, validate, test=p.preprocess_data(dataset,'target_name','list_of_specific_columns')
>>> preprocessed_dataset.to_excel("df_feature.xlsx")


Feature Engineering
^^^^^^^^^^^^^^

>>> from easycheml.modelling import feature_engineering as f
>>> feature=f.feature_thru_correlation('df_feature.xlsx', 'target_name', 0.4, 'pearson')



Data Modelling
^^^^^^^^^^^^^^^

>>> from easycheml.modelling import Regressors 
>>> model=Regressors('df_feature.xlsx','target_name',0.6,0.2)
>>> model.ensemble_models("RF", None)      # Random Forest Regressor 
>>> model.ensemble_models("AdaBoost",None) # AdaBoost Regressor

>>> # Hyperparameter for tuning above Random Forest Regressor
>>> parameters = {
    'n_estimators' :[50,100,200,300,400,500,600,700,800,900,1000],
    'criterion' : ["squared_error", "friedman_mse", "absolute_error"],
    'max_depth' : [3,5,7,9,11,13,15,17,19,21,23,25,27,29,31],
    'min_samples_split' : [5,10,20,30,40,50,60,70,80,90,100],
    'bootstrap':[True],
    'min_samples_leaf':[5,10,20,30,40,50,60,70,80,90,100],
                }
>>> model.ensemble_models("RF",parameters)

>>> # Deep Learning Sequential Model
>>> num_max_trials=3
>>> num_executions_per_trial=3
>>> num_epochs=10
>>> num_batch_size=32
>>> model.dnn_sequential_model(num_max_trials,num_executions_per_trial,num_epochs,num_batch_size)


Data Postprocessing and Visualization
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^




Installation
------------

First, install `Tensorflow <https://www.tensorflow.org/install>`_. Then, install EasyCheml with

>>> pip install easycheml

