Fastreport

Get report of different metrices for classification and regression problem for many popular algorithms with single line of code. You have to pass only features(dataframe) and target(series) as arguments

Link to PyPI

Link to Classification detailed example

Link to Regression detailed example

Installation

Run the following to install:

pip install fastreport

Usage

Classification

import report


report.report_classification(df_features,df_target,algorithms='default',test_size=0.3,scaling=None,
                             large_data=False,encode='dummy',average='binary',change_data_type = False,
                             threshold=8,random_state=None):
   

parameters
----------------------------

df_features : Pandas DataFrame

df_target : Pandas Series

algorithms : List ,'default'=
             [LogisticRegression(),
             GaussianNB(),
             DecisionTreeClassifier(),
             RandomForestClassifier(),
             GradientBoostingClassifier(),
             AdaBoostClassifier()]
             The above are the default algorithms, if one needs any specific algorithms, they have to import
             libraries then pass the instances of alogorith as list
             For example, if one needs random forest and adaboost only, then pass 
             
             algorithms=[RandomForestClassifier(max_depth=8),AdaBoostClassifier()]
             But, these libraries must be imported before passing into above list like
             

test_size: If float, should be between 0.0 and 1.0 and represent the proportion of the 
           dataset to include in the test split.

scaling : {'Standard-scalar', 'Min-Max'} or None , default=None

encode : {'dummy','onehot','label'} ,default='dummy'

change_data_type : bool, default=False
                   Some columns will be of numerical datatype though there are only 2-3 unique values in that column,
                   so these columns must be converted to object as it is more relevant.
                   By setting change_data_type= True , these columns will be converted into object datatype

threshold : int ,default=8
            Maximum unique value a column can have

large_data : bool, default=False
            If the dataset is large then the parameter large_data should be set to True, 
            make sure if your system has enough memory before setting Large_data=True

            
average : {'micro', 'macro', 'samples','weighted', 'binary'} or None, default='binary'
This parameter is required for multiclass/multilabel targets.
If ``None``, the scores for each class are returned. Otherwise, this
determines the type of averaging performed on the data:

``'binary'``:
    Only report results for the class specified by ``pos_label``.
    This is applicable only if targets (``y_{true,pred}``) are binary.
``'micro'``:
    Calculate metrics globally by counting the total true positives,
    false negatives and false positives.
``'macro'``:
    Calculate metrics for each label, and find their unweighted
    mean.  This does not take label imbalance into account.
``'weighted'``:
    Calculate metrics for each label, and find their average weighted
    by support (the number of true instances for each label). This
    alters 'macro' to account for label imbalance; it can result in an
    F-score that is not between precision and recall.
``'samples'``:
    Calculate metrics for each instance, and find their average (only
    meaningful for multilabel classification where this differs from
    :func:`accuracy_score`).
    
random_state : int, RandomState instance or None, default=None

Regression

import report

report.report_regression(df_features,df_target,algorithms='default',test_size=0.3,
                      scaling=None,large_data=False,change_data_type=True,encode='dummy',
                      threshold=8,random_state=None):

parameters
----------------------------

df_features : Pandas DataFrame

df_target : Pandas Series

 algorithms : List ,'default'=
             [LinearRegression(),
             Lasso(),
             Ridge(),
             RandomForestRegressor(),
             GradientBoostingRegressor(),
             AdaBoostRegressor()]
             The above are the default algorithms, if one needs any specific algorithms, they have to import
             libraries then pass the instances of alogorith as list
             For example, if one needs random forest and adaboost only, then pass 
             
             algorithms=[RandomForestRegressor(max_depth=8),AdaBoostRegressor()]
             But, these libraries must be imported before passing into above list like
             
test_size: If float, should be between 0.0 and 1.0 and represent the proportion of the 
           dataset to include in the test split.

scaling : {'Standard-scalar', 'Min-Max'} or None , default=None

encode : {'dummy','onehot','label'} ,default='dummy'

change_data_type : bool, default=False
                   Some columns will be of numerical datatype though there are only 2-3 unique values in that column,
                   so these columns must be converted to object as it is more relevant.
                   By setting change_data_type= True , these columns will be converted into object datatype

threshold : int ,default=8
            Maximum unique value a column can have
            
large_data : bool, default=False
            If the dataset is large then the parameter large_data should be set to True, 
            make sure if your system has enough memory before setting Large_data=True
            
random_state : int, RandomState instance or None, default=None

Future works

1. Optimization
2. Add more functionality

Drawbacks

1. Not suitable for very large datasets
2. Limited to existing users only

License

© 2021 KISHORE S This repository is licensed under the MIT license. See LICENSE for details.