basicanalysis 0.1.1

A quick way to see the best supervised learning method for your dataset or best configuration for the chosen method.

Performance Overview of Supervised Learning methods

Do not know which supervised learning method is good for your dataset? Would you like to know it in just few seconds?

Congratulations:

You are about to learn about a package which gives you the solution to all above problems!

This small package of merely few bytes and code written in less than 100 lines, provide you the overview of all fundamental metrics measured for almost all supervised learning method.

Models evaluated:	Metrics considered:
Decision Trees	Logistic Regression	Accuracy	Precision
Naive Bayes	SVM	Jaccard Score	F1_Score
Neural Networks	K-NN	R (Corr. Coeff.)	ROC AUC
Random Forest	Adaboost	MSE	Log Loss

---

Mandatory inputs required:

A Pandas DataFrame

Optional inputs in the given order:

• Column numbers for the predictors in the form of a LIST

  Default: It will take all columns except the last one.

• Column number for the response in the form of a LIST

  Default: It will take the last column.

• Test size in Float Ex. 0.3 for 30% Test Size.

  Default: 0.25 (25% Test size) will be assumed.

---

How to install:

Type pip install basicanalysis in command line to install the package

To call the module from this package, type from basicanalysis.basicanalysis import basicanalysis

Note : To import other modules, use the following command.

from basicanalysis.basicanalysis import knn from basicanalysis.basicanalysis import knn_10fold from basicanalysis.basicanalysis import *

---

Neat! Isn’t it?

MAJOR UPDATE: 0.0.3 -> 0.1.0, BasicAnalysis -> basicanalysis

• Added class knn to run K-NN method to compare with multiple inputs

• Added class knn_10fold to run K-nn method on training data with 10-fold cross validation, comparing multiple inputs.

---

README file for the task

Written in reStructuredText or .rst file, and used to generate the project page on PyPI. Images coming soon…