klib 0.1.3

Customized data preprocessing functions for frequent tasks.

klib

klib is a Python library for importing, cleaning, analyzing and preprocessing data. While the focus is on these steps, future versions will include modules and functions for model creation and optimization to provide more of an end-to-end solution. Explanations on key functionalities can be found on Medium / TowardsDataScience or in the examples section.

Installation

Use the package manager pip to install klib.

pip install klib
pip install --upgrade klib

Alternatively, to install this package with conda run:

conda install -c conda-forge klib

Usage

import klib
import pandas as pd

df = pd.DataFrame(data)

# klib.describe functions for visualizing datasets
- klib.cat_plot(df) # returns a visualization of the number and frequency of categorical features
- klib.corr_mat(df) # returns a color-encoded correlation matrix
- klib.corr_plot(df) # returns a color-encoded heatmap, ideal for correlations
- klib.dist_plot(df) # returns a distribution plot for every numeric feature
- klib.missingval_plot(df) # returns a figure containing information about missing values

# klib.clean functions for cleaning datasets
- klib.data_cleaning(df) # performs datacleaning (drop duplicates & empty rows/cols, adjust dtypes,...)
- klib.clean_column_names(df) # cleans and standardizes column names, also called inside data_cleaning()
- klib.convert_datatypes(df) # converts existing to more efficient dtypes, also called inside data_cleaning()
- klib.drop_missing(df) # drops missing values, also called in data_cleaning()
- klib.mv_col_handling(df) # drops features with high ratio of missing vals based on informational content
- klib.pool_duplicate_subsets(df) # pools subset of cols based on duplicates with min. loss of information

# klib.preprocess functions for data preprocessing (feature selection, scaling, ...)
- klib.train_dev_test_split(df) # splits a dataset and a label into train, optionally dev and test sets
- klib.feature_selection_pipe() # provides common operations for feature selection
- klib.num_pipe() # provides common operations for preprocessing of numerical data
- klib.cat_pipe() # provides common operations for preprocessing of categorical data
- klib.preprocess.ColumnSelector() # selects num or cat columns, ideal for a Feature Union or Pipeline
- klib.preprocess.PipeInfo() # prints out the shape of the data at the specified step of a Pipeline

Examples

Find all available examples as well as applications of the functions in klib.clean() with detailed descriptions here.

klib.missingval_plot(df) # default representation of missing values in a DataFrame, plenty of settings are available

klib.corr_plot(df, split='pos') # displaying only positive correlations, other settings include threshold, cmap...
klib.corr_plot(df, split='neg') # displaying only negative correlations

klib.corr_plot(df, target='wine') # default representation of correlations with the feature column

klib.dist_plot(df) # default representation of a distribution plot, other settings include fill_range, histogram, ...

klib.cat_plot(data, top=4, bottom=4) # representation of the 4 most & least common values in each categorical column

Further examples, as well as applications of the functions in klib.clean() can be found here.

Contributing

Pull requests and ideas, especially for further functions are welcome. For major changes or feedback, please open an issue first to discuss what you would like to change.

License

MIT