Added dataframe support and correlation plot

README.md

@@ -1,2 +1,40 @@
 # PyCorr
-Simple package to calculate correlation between variables. Currently provides correlation between nominal variables. Correlation between mixed variables would be added soon.
+A simple library to calculate correlation between variables. Currently provides correlation between nominal variables.
+
+Based on statistical methodology like Cramer'V and Tschuprow'T allows to gauge the correlation between categorical variables. Ability to plot the correlation in form of heatmap is also provided.
+
+## Usage example
+
+
+## Development setup
+
+## Release History
+
+* 0.1.3
+    * CHANGE: Remove `setDefaultXYZ()`
+    * ADD: Ability to pass dataframe to get correlation matrix
+    * ADD: Ability to plot the correlation in form of heatmap
+* 0.1.2
+    * Added as first release
+* 0.1.1
+    * Test release
+
+## Meta
+
+Anurag Kumar Mishra – [https://github.com/anuragithub](https://github.com/anuragithub) – anuragkm25@outlook.com
+
+Distributed under the GNU license. See ``LICENSE`` for more information.
+
+[https://github.com/MavericksDS/pycorr](https://github.com/MavericksDS/pycorr)
+
+
+## Contributing
+
+1. Fork it (<https://github.com/MavericksDS/pycorr>)
+2. Create your feature branch (`git checkout -b feature/fooBar`)
+3. Commit your changes (`git commit -am 'Add some fooBar'`)
+4. Push to the branch (`git push origin feature/fooBar`)
+5. Create a new Pull Request
+
+<!-- Markdown link & img dfn's -->
+[pip-url]: https://pypi.org/project/pycorr/

pycorrcat/pycorrcat.py

@@ -2,6 +2,9 @@
 import pandas as pd
 import scipy.stats as stats
 import numpy as np
+from typing import List
+import matplotlib as matplotlib
+import seaborn as sns
 
 def fillna(object):
     if isinstance(object, pd.Series):
@@ -62,4 +65,82 @@ def corr(x,
         return corr_coeff
     except:
         warnings.warn("Error calculating Cramer's V",RuntimeWarning)
-        return corr_coeff
+        return corr_coeff
+
+
+def corr_matrix(data: pd.DataFrame, 
+                columns: List,
+                bias_correction: bool=True,
+                Tschuprow: bool=False) -> pd.DataFrame:
+    """
+    Calculates correlation for all the columns provided and returns pandas like correlation matrix.
+    The two measures supported are:
+    1. Cramer'V ( default )
+    2. Tschuprow'T
+
+    Parameters:
+    -----------
+    data : pandas DataFrame
+        A pandas DataFrame containing the categorical columns
+    columns : list 
+        A list of categorical columns
+    bias_correction : Boolean, default = True
+    Tschuprow : Boolean, default = False
+               For choosing Tschuprow as measure
+    Returns:
+    --------
+    pandas dataframe object similar to pandas.DataFrame.corr()
+    """
+    # checking length of columns
+    if not columns.__len__()>0 or set(data.columns.values).intersection(columns).__len__()>0 :
+        ValueError("Check the columns list provided")
+
+    target_data = data.filter(columns)
+    cols = target_data.columns.values
+    shape = target_data.columns.__len__()
+
+    matrix = np.zeros((shape, shape))
+    for x,i in enumerate(cols):
+        temp = np.zeros((0, shape))
+        for j in cols:
+            temp = np.append(temp,corr(target_data[i], target_data[j], bias_correction=bias_correction, Tschuprow=Tschuprow))
+        matrix[x] = temp
+
+    corr_matrix = pd.DataFrame(data=matrix,
+                            index=cols,
+                            columns=cols)
+    return corr_matrix
+
+def plot_corr(data: pd.DataFrame,
+              columns: List,
+              diagonal: str = False,
+              bias_correction: bool=True,
+              Tschuprow: bool=False
+            ) -> matplotlib.axes.Axes:
+    """
+    Plots correlation matrix for all the columns provided and returns Matplotlib axes.
+    The two measures supported are:
+    1. Cramer'V ( default )
+    2. Tschuprow'T
+
+    Parameters:
+    -----------
+    data : pandas DataFrame
+        A pandas DataFrame containing the categorical columns
+    columns : list 
+        A list of categorical columns
+    diagonal :  string
+        When true gives a masked version of heatmap
+    bias_correction : Boolean, default = True
+    Tschuprow : Boolean, default = False
+               For choosing Tschuprow as measure
+    Returns:
+    --------
+    ax : matplotlib Axes
+    Axes object with the heatmap.
+    """
+    corr = corr_matrix(data, columns, bias_correction=bias_correction, Tschuprow=Tschuprow)
+    if(diagonal):
+        mask = np.triu(corr)
+        return sns.heatmap(corr, annot=True, mask=mask)
+    return sns.heatmap(corr, annot=True)

requirements.txt

@@ -1,9 +1,12 @@
 bleach==3.1.5
 certifi==2020.4.5.1
 chardet==3.0.4
+cycler==0.10.0
 docutils==0.16
 idna==2.9
 keyring==21.2.1
+kiwisolver==1.2.0
+matplotlib==3.2.1
 numpy==1.18.4
 packaging==20.3
 pandas==1.0.3
@@ -17,9 +20,9 @@ readme-renderer==26.0
 requests==2.23.0
 requests-toolbelt==0.9.1
 scipy==1.4.1
+seaborn==0.10.1
 six==1.14.0
 tqdm==4.46.0
 twine==3.1.1
 urllib3==1.25.9
 webencodings==0.5.1
-wheel==0.33.1

setup.py

@@ -10,7 +10,7 @@
 
 setup(
     name='pycorr',
-    version='0.1.2',
+    version='0.1.3',
     description='Python package for calculating correlation amongst categorical variables',
     long_description_content_type='text/markdown',
     long_description=long_description,