data_visualization

# This is program is only for reference for various plotting methods

# Import Standard Libraries 
import numpy as np
import pandas as pd
from numpy.random import randn
import warnings
warnings.filterwarnings('ignore')

# stats libraries
from scipy import stats

# Plotting libraries
import matplotlib as mpl
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline

url = 'http://en.wikipedia.org/wiki/Histogram'
dataset1 = randn(100)
# check dataset
plt.hist(dataset1)

dataset2 = randn(80)
plt.hist(dataset2,color='indianred')

plt.hist(dataset1,normed=True,color='indianred',alpha=0.5,bins=20)
plt.hist(dataset2,normed=True,alpha=0.5,bins=20)

# Try Joint plot
data1 = randn(1000)
data2 = randn(1000)
sns.jointplot(data1,data2)
sns.jointplot(data1,data2,kind='hex')  # with hex

#Kernel Density Estimation Plots (Kde) For more details refer KDE wikipedia (Gaussian basis function)
dataset = randn(25)
sns.rugplot(dataset)
plt.ylim(0,1)

plt.hist(dataset,alpha=0.3)
sns.rugplot(dataset)

sns.rugplot(dataset)

x_min = dataset.min() - 2
x_max = dataset.max() + 2

x_axis = np.linspace(x_min,x_max,100)

bandwidth = ( (4*dataset.std()**5) / (3*len(dataset))) ** 0.2

kernel_list = []

for data_point in dataset:
    # Create a kernel for each point and append it to kernel_list
    kernel = stats.norm(data_point,bandwidth).pdf(x_axis)
    kernel_list.append(kernel)
    
    # Scale for plotting
    kernel = kernel / kernel.max()
    kernel = kernel * 0.4
    
    plt.plot(x_axis,kernel,color='grey',alpha=0.5)

plt.ylim(0,1)

sum_of_kde = np.sum(kernel_list,axis=0)

fig = plt.plot(x_axis,sum_of_kde,color='indianred')

sns.rugplot(dataset)

plt.yticks([])

plt.suptitle("Sum of basis functions")

sns.rugplot(dataset,color='black')

for bw in np.arange(0.5,2,0.25):
    sns.kdeplot(dataset,bw=bw,lw=1.8,label=bw)

# Read url
url = 'http://en.wikipedia.org/wiki/Kernel_(statistics)'
kernel_options = ['biw','cos','epa','gau','tri','triw']

for kern in kernel_options:
    sns.kdeplot(dataset,kernel=kern,label=kern,shade=True)
    
sns.kdeplot(dataset,vertical=True)

url = 'http://en.wikipedia.org/wiki/Cumulative_distribution_function'
sns.kdeplot(dataset,cumulative = True)

mean = [0,0]
cov = [[1,0],[0,100]]
dataset2 = np.random.multivariate_normal(mean,cov,1000)
dframe = pd.DataFrame(dataset2,columns=['X','Y'])
sns.kdeplot(dframe)

sns.kdeplot(dframe.X,dframe.Y,shade=True)
sns.kdeplot(dframe,bw=1)
sns.kdeplot(dframe,bw='silverman')

sns.jointplot('X','Y',dframe,kind='kde')

# Combining Plot Styles
dataset = randn(100)
sns.distplot(dataset,bins=25)
sns.distplot(dataset,bins=25,rug=True,hist=False)

sns.distplot(dataset,bins=25,
             kde_kws={'color':'indianred','label':'KDE PLOT'},
             hist_kws={'color':'blue','label':'HIST'})
           
from pandas import Series
ser1 = Series(dataset,name='My_data')
ser1

sns.distplot(ser1,bins=25)
# Box and Violin Plots
url = 'http://en.wikipedia.org/wiki/Box_plot#mediaviewer/File:Boxplot_vs_PDF.svg'

data1 = randn(100)
data2 = randn(100)
sns.boxplot([data1,data2])

sns.boxplot([data1,data2],whis=np.inf)
# Normal Dist
data1 = stats.norm(0,5).rvs(100)

# Two gamma dist. Concatenated together
data2 = np.concatenate([stats.gamma(5).rvs(50)-1,
                       -1*stats.gamma(5).rvs(50)])

# Box Plot both data1 and data2
sns.boxplot([data1,data2],whis=np.inf)
sns.violinplot(data1,data2)

sns.violinplot(data2,bw=0.01)
sns.violinplot(data1,inner='stick')

# Regression Plots
tips = sns.load_dataset('tips')
tips.head()

sns.lmplot('total_bill','tip',tips)
sns.lmplot('total_bill','tip',tips,
           scatter_kws={'marker':'o','color':'indianred'},
           line_kws={'linewidth':1,'color':'blue'})
           
sns.lmplot('total_bill','tip',tips,order=4,
           scatter_kws={'marker':'o','color':'indianred'},
           line_kws={'linewidth':1,'color':'blue'})
           
sns.lmplot('total_bill','tip',tips,fit_reg=False)

tips['tip_pect']=100*(tips['tip']/tips['total_bill'])
tips.head()

sns.lmplot('size','tip_pect',tips)

url = 'http://en.wikipedia.org/wiki/jitter'
sns.lmplot('size','tip_pect',tips,x_jitter=.1)

sns.lmplot('size','tip_pect',tips,x_estimator=np.mean)
sns.lmplot('total_bill','tip_pect',tips,hue='sex',markers=['x','o'])

sns.lmplot('total_bill','tip_pect',tips,hue='day')

url = 'http://en.wikipedia.org/wiki/Local_regression'

sns.lmplot('total_bill','tip_pect',tips,lowess=True,line_kws={'color':'black'})
sns.regplot('total_bill','tip_pect',tips)

fig, (axis1,axis2) = plt.subplots(1,2,sharey=True)

sns.regplot('total_bill','tip_pect',tips,ax=axis1)
sns.violinplot(tips['tip_pect'],tips['size'],color='indianred',ax=axis2)

# Heatmap and clustered Matrics
flight_dframe = sns.load_dataset('flights')
flight_dframe.head()

flight_dframe = flight_dframe.pivot('month','year','passengers')
flight_dframe

sns.heatmap(flight_dframe)
sns.heatmap(flight_dframe,annot=True,fmt='d')

sns.heatmap(flight_dframe,center=flight_dframe.loc['January',1955])

f,(axis1,axis2) = plt.subplots(2,1)
yearly_flights = flight_dframe.sum()
years = pd.Series(yearly_flights.index.values)
years = pd.DataFrame(years)
flights = pd.Series(yearly_flights.values)
flights = pd.DataFrame(flights)
year_dframe = pd.concat((years,flights),axis=1)
year_dframe.columns = ['Year','Flights']
sns.barplot('Year',y='Flights',data=year_dframe,ax=axis1)
sns.heatmap(flight_dframe,cmap='Blues',ax=axis2,cbar_kws={'orientation':'horizontal'})

sns.clustermap(flight_dframe)

sns.clustermap(flight_dframe,col_cluster=False)
sns.clustermap(flight_dframe,standard_scale=1)

sns.clustermap(flight_dframe,standard_scale=0)

sns.clustermap(flight_dframe,z_score=1)