# -*- coding: utf-8 -*-
"""
Created on Mon Nov 30 13:33:31 2020

Task: Statistical analysis and visualization of Iris dataset
using pandas and seaborn

Python tools    
Libraries: numpy, sklearn, pandas, seaborn
Modules: pyplot, plotting
Classes:  
Functions: 

@author: Márton Ispány
"""

import numpy as np;  # importing numerical computing package
from matplotlib import pyplot as plt;  # importing MATLAB-like plotting framework
import pandas as pd;  # importing pandas data analysis tool
import seaborn as sns;  # importing statistical data visualization
from sklearn.datasets import load_iris; # importing Iris dataset

# Loading dataset as frame
iris = load_iris(as_frame=True);
n = iris.data.shape[0]; # number of records
p = iris.data.shape[1]; # number of attributes
k = iris.target_names.shape[0];

iris_by_target = iris.frame.groupby(by='target');  # grouping by target
# Exporting a target group from the grouped dataframe
for i in range(k):
    if iris.target_names[i]=='setosa': setosa_ind=i;
iris_setosa = iris_by_target.get_group(setosa_ind);
first = 5;
# Printing the first observations by target
for target_value, group in iris_by_target:
   print ('Target: ',iris.target_names[target_value]);
   print (group[0:first]);
 
# Basic descriptive stats
mean_by_target = iris_by_target.mean();  # mean
std_by_target = iris_by_target.std();  # standard deviation
corr_by_target = iris_by_target.corr();  #  correlations
desc_stat_by_target = iris_by_target.describe();  # desc stat with quantiles

# Printing the results
print(mean_by_target);

# Plotting using multidimensional tools of pandas
# Andrews curves
plt.figure(1);
pd.plotting.andrews_curves(iris.frame,'target',color=['blue','green','red']);
plt.show();
# Parallel axis
plt.figure(2);
pd.plotting.parallel_coordinates(iris.frame,'target',color=['blue','green','red']);
plt.show();

# Scatter matrix
pd.plotting.scatter_matrix(iris.data);
                         

# Loading seaborn's default theme and color palette
sns.set(); 

# Plotting two attributes
# Default axis
x_axis = 0;  # x axis attribute (0..3)
y_axis = 1;  # y axis attribute (0..3)
# Enter axis from consol
user_input = input('X axis [0..3, default:0]: ');
if len(user_input) != 0 and np.int8(user_input)>=0 and np.int8(user_input)<=3 :
    x_axis = np.int8(user_input);
user_input = input('Y axis [0..3, default:1]: ');
if len(user_input) != 0 and np.int8(user_input)>=0 and np.int8(user_input)<=3 :
    y_axis = np.int8(user_input); 
sns.scatterplot(data=iris.frame, x=iris.feature_names[x_axis], 
                        y=iris.feature_names[y_axis], hue='target');
sns.histplot(data=iris.frame, x=iris.feature_names[x_axis], 
                        y=iris.feature_names[y_axis], hue='target');
plt.figure();
sns.histplot(data=iris.frame, x=iris.feature_names[x_axis], hue='target');
plt.show();

sns.boxplot(data=iris.frame);

plt.figure();
sns.pairplot(data=iris.frame,hue='target');
plt.show();