14 - Intermediate Data Visualization with Seaborn

                            //^//^//^//^//^//^//^//^//^ Seaborn Introduction ^\\^\\^\\^\\^\\^\\^\\^\\^\\
                            
Reading a csv file
--------------------------------------------------
# import all modules
import pandas as pd
import seaborn as sb
import matplotlib.pyplot as plt

# Read in the DataFrame
df = pd.read_csv(grant_file)


Comparing a histogram and displot
 --------------------------------------------------
# Display pandas histogram
df['Award_Amount'].plot.hist()
plt.show()

# Clear out the pandas histogram
plt.clf()
---------
# Display a Seaborn distplot
sns.displot(df['Award_Amount'])
plt.show()

# Clear the distplot
plt.clf()


Plot a histogram
 --------------------------------------------------
 # Create a distplot
sns.displot(df['Award_Amount'],
             kde=False,
             bins=20)

# Display the plot
plt.show()


Rug plot and kde shading
 --------------------------------------------------
 # Create a distplot of the Award Amount
sns.displot(df['Award_Amount'],kind = 'kde', fill = True,
             rug=True)

# Plot the results
plt.show()


Create a regression plot
 --------------------------------------------------
 # Create a regression plot of premiums vs. insurance_losses
sns.regplot(x="insurance_losses", y="premiums", data=df)


# Display the plot
plt.show()
---------
# Create an lmplot of premiums vs. insurance_losses
sns.lmplot(x="insurance_losses", y="premiums", data=df)



# Display the second plot
plt.show()



Plotting multiple variables
 --------------------------------------------------
# Create a regression plot using hue
sns.lmplot(data=df,
           x="insurance_losses",
           y="premiums",
           hue="Region")

# Show the results
plt.show()


Facetting multiple regressions
 --------------------------------------------------
 # Create a regression plot with multiple rows
sns.lmplot(data=df,
           x="insurance_losses",
           y="premiums",
           hue="Region",row='Region')

# Show the plot
plt.show()


                            //^//^//^//^//^//^//^//^//^ Customizing Seaborn Plots ^\\^\\^\\^\\^\\^\\^\\^\\^\\
                           
Setting the default style
 --------------------------------------------------
# Plot the pandas histogram
df['fmr_2'].plot.hist()
plt.show()
plt.clf()

# Set the default seaborn style
sns.set()

# Plot the pandas histogram again
df['fmr_2'].plot.hist()
plt.show()
plt.clf()


Comparing styles
 --------------------------------------------------
sns.displot(df['fmr_2'])
sns.set_style('dark')
plt.show()
plt.clf()
---------
sns.displot(df['fmr_2'])
sns.set_style('whitegrid')
plt.show()
plt.clf()


Removing spines
--------------------------------------------------
# Set the style to white
sns.set_style('white')

# Create a regression plot
sns.lmplot(data=df,
           x='pop2010',
           y='fmr_2')

# Remove the spines
sns.despine()

# Show the plot and clear the figure
plt.show()
plt.clf()


Matplotlib color codes
--------------------------------------------------
# Set style, enable color code, and create a magenta displot
sns.set(color_codes = True)
sns.displot(df['fmr_3'], color='m')

# Show the plot
plt.show()


Using default palettes
--------------------------------------------------
# Loop through differences between bright and colorblind palettes
for p in ['bright', 'colorblind']:
    sns.set_palette(p)
    sns.displot(df['fmr_3'])
    plt.show()
    
    # Clear the plots    
    plt.clf()
    
    
Creating Custom Palettes
--------------------------------------------------
sns.palplot(sns.color_palette("Purples", 8))
plt.show()
---------
sns.palplot(sns.color_palette('husl', 10))
plt.show()
---------
sns.palplot(sns.color_palette('coolwarm', 6))
plt.show()


Using matplotlib axes
--------------------------------------------------
# Create a figure and axes
fig, ax = plt.subplots()

# Plot the distribution of data
sns.histplot(df['fmr_3'], ax=ax)

# Create a more descriptive x axis label
ax.set(xlabel="3 Bedroom Fair Market Rent")

# Show the plot
plt.show()


Additional plot customizations
--------------------------------------------------
# Create a figure and axes
fig, ax = plt.subplots()

# Plot the distribution of 1 bedroom rents
sns.histplot(df['fmr_1'], ax=ax)

# Modify the properties of the plot
ax.set(xlabel="1 Bedroom Fair Market Rent",
       xlim=(100,1500),
       label="US Rent")

# Display the plot
plt.show()


Adding annotations
--------------------------------------------------
# Create a figure and axes. Then plot the data
fig, ax = plt.subplots()
sns.histplot(df['fmr_1'], ax=ax)

# Customize the labels and limits
ax.set(xlabel="1 Bedroom Fair Market Rent", xlim=(100,1500), title="US Rent")

# Add vertical lines for the median and mean
ax.axvline(x=df['fmr_1'].median(), color='m', label='Median', linestyle='--', linewidth=2)
ax.axvline(x=df['fmr_1'].mean(), color='b', label='Mean', linestyle='-', linewidth=2)

# Show the legend and plot the data
ax.legend()
plt.show()


Multiple plots
--------------------------------------------------
# Create a plot with 1 row and 2 columns that share the y axis label
fig, (ax0, ax1) = plt.subplots(nrows=1, ncols=2, sharey=True)

# Plot the distribution of 1 bedroom apartments on ax0
sns.histplot(df['fmr_1'], ax=ax0)
ax0.set(xlabel="1 Bedroom Fair Market Rent", xlim=(100,1500))

# Plot the distribution of 2 bedroom apartments on ax1
sns.histplot(df['fmr_2'], ax=ax1)
ax1.set(xlabel="2 Bedroom Fair Market Rent", xlim=(100,1500))

# Display the plot
plt.show()


                            //^//^//^//^//^//^//^//^//^ Additional Plot Types ^\\^\\^\\^\\^\\^\\^\\^\\^\\
                            
stripplot() and swarmplot()
--------------------------------------------------
# Create the stripplot
sns.stripplot(data=df,
         x='Award_Amount',
         y='Model Selected',
         jitter=True)
plt.show()
---------
# Create and display a swarmplot with hue set to the Region
sns.swarmplot(data=df,
         x='Award_Amount',
         y='Model Selected',
         hue='Region')

plt.show()


boxplots, violinplots and boxenplots
--------------------------------------------------
# Create a boxplot
sns.boxplot(data=df,
         x='Award_Amount',
         y='Model Selected')

plt.show()
plt.clf()
---------
# Create a violinplot with the husl palette
sns.violinplot(data=df,
         x='Award_Amount',
         y='Model Selected',
         palette='husl')

plt.show()
plt.clf()
---------
# Create a boxenplot with the Paired palette and the Region column as the hue
sns.boxenplot(data=df,
         x='Award_Amount',
         y='Model Selected',
         palette='Paired',
         hue='Region')

plt.show()
plt.clf()


barplots, pointplots and countplots
--------------------------------------------------
# Show a countplot with the number of models used with each region a different color
sns.countplot(data=df,
         y="Model Selected",
         hue="Region")

plt.show()
plt.clf()
---------
# Create a pointplot and include the capsize in order to show caps on the error bars
sns.pointplot(data=df,
         y='Award_Amount',
         x='Model Selected',
         capsize=.1)

plt.show()
plt.clf()
---------
# Create a barplot with each Region shown as a different color
sns.barplot(data=df,
         y='Award_Amount',
         x='Model Selected',
         hue='Region')

plt.show()
plt.clf()


Regression and residual plots
--------------------------------------------------
# Display a regression plot for Tuition
sns.regplot(data=df,
         y='Tuition',
         x='SAT_AVG_ALL',
         marker='^',
         color='g')

plt.show()
plt.clf()
---------
# Display the residual plot
sns.residplot(data=df,
          y='Tuition',
          x='SAT_AVG_ALL',
          color='g')

plt.show()
plt.clf()


Regression plot parameters
--------------------------------------------------
# Plot a regression plot of Tuition and the Percentage of Pell Grants
sns.regplot(data=df,
            y='Tuition',
            x='PCTPELL')

plt.show()
plt.clf()
---------
# Create another plot that estimates the tuition by PCTPELL
sns.regplot(data=df,
            y='Tuition',
            x='PCTPELL',
            x_bins=5)

plt.show()
plt.clf()
---------
# The final plot should include a line using a 2nd order polynomial
sns.regplot(data=df,
            y='Tuition',
            x='PCTPELL',
            x_bins=5,
            order=2)

plt.show()
plt.clf()


Binning data
--------------------------------------------------
# Create a scatter plot by disabling the regression line
sns.regplot(data=df,
            y='Tuition',
            x='UG',
            fit_reg=False)

plt.show()
plt.clf()
---------
# Create a scatter plot and bin the data into 5 bins
sns.regplot(data=df,
            y='Tuition',
            x='UG',
            x_bins=5)

plt.show()
plt.clf()
---------
# Create a regplot and bin the data into 8 bins
sns.regplot(data=df,
         y='Tuition',
         x='UG',
         x_bins=8)

plt.show()
plt.clf()


Creating heatmaps
--------------------------------------------------
# Create a crosstab table of the data
pd_crosstab = pd.crosstab(df["Group"], df["YEAR"])
print(pd_crosstab)

# Plot a heatmap of the table
sns.heatmap(pd_crosstab)

# Rotate tick marks for visibility
plt.yticks(rotation=0)
plt.xticks(rotation=90)

plt.show()


Customizing heatmaps
--------------------------------------------------
# Create the crosstab DataFrame
pd_crosstab = pd.crosstab(df["Group"], df["YEAR"])

# Plot a heatmap of the table with no color bar and using the BuGn palette
sns.heatmap(pd_crosstab, cbar=False, cmap="BuGn", linewidths=0.3)

# Rotate tick marks for visibility
plt.yticks(rotation=0)
plt.xticks(rotation=90)

#Show the plot
plt.show()
plt.clf()


                            //^//^//^//^//^//^//^//^//^ Creating Plots on Data Aware Grids ^\\^\\^\\^\\^\\^\\^\\^\\^\\
                            
Building a FacetGrid             
--------------------------------------------------
# Create FacetGrid with Degree_Type and specify the order of the rows using row_order
g2 = sns.FacetGrid(df, 
             col="Degree_Type",row = 'Degree_Type',
             row_order=['Graduate', 'Bachelors', 'Associates', 'Certificate'])

# Map a pointplot of SAT_AVG_ALL onto the grid
g2.map(sns.pointplot, 'SAT_AVG_ALL')

# Show the plot
plt.show()
plt.clf()


Using a catplot
--------------------------------------------------
# Create a factor plot that contains boxplots of Tuition values
sns.catplot(data=df,
         x='Tuition',
         kind='box',
         row='Degree_Type')

plt.show()
plt.clf()
---------
# Create a facetted pointplot of Average SAT_AVG_ALL scores facetted by Degree Type 
sns.catplot(data=df,
        x='SAT_AVG_ALL',
        kind='point',
        row='Degree_Type',
        row_order=['Graduate', 'Bachelors', 'Associates', 'Certificate'])

plt.show()
plt.clf()


Using a lmplot
--------------------------------------------------
# Create a FacetGrid varying by column and columns ordered with the degree_order variable
g = sns.FacetGrid(df, col="Degree_Type", col_order=degree_ord)

# Map a scatter plot of Undergrad Population compared to PCTPELL
g.map(plt.scatter, 'UG', 'PCTPELL')

plt.show()
plt.clf()
---------
# Re-create the previous plot as an lmplot
sns.lmplot(data=df,
        x='UG',
        y='PCTPELL',
        col="Degree_Type",
        col_order=degree_ord)

plt.show()
plt.clf()
---------
# Create an lmplot that has a column for Ownership, a row for Degree_Type and hue based on the WOMENONLY column
sns.lmplot(data=df,
        x='SAT_AVG_ALL',
        y='Tuition',
        col="Ownership",
        row='Degree_Type',
        row_order=['Graduate', 'Bachelors'],
        hue='WOMENONLY',
        col_order=inst_ord)

plt.show()
plt.clf()


Building a PairGrid
--------------------------------------------------
# Create a PairGrid with a scatter plot for fatal_collisions and premiums
g = sns.PairGrid(df, vars=["fatal_collisions", "premiums"])
g2 = g.map(sns.scatterplot)

plt.show()
plt.clf()
---------
# Create the same PairGrid but map a histogram on the diag
g = sns.PairGrid(df, vars=["fatal_collisions", "premiums"])
g2 = g.map_diag(sns.histplot)
g3 = g2.map_offdiag(sns.scatterplot)

plt.show()
plt.clf()


Using a pairplot
--------------------------------------------------
# Create a pairwise plot of the variables using a scatter plot
sns.pairplot(data=df,
        vars=["fatal_collisions", "premiums"],
        kind='scatter')

plt.show()
plt.clf()
---------
# Plot the same data but use a different color palette and color code by Region
sns.pairplot(data=df,
        vars=["fatal_collisions", "premiums"],
        kind='scatter',
        hue='Region',
        palette='RdBu',
        diag_kws={'alpha':.5})

plt.show()
plt.clf()


Additional pairplots
--------------------------------------------------
# Build a pairplot with different x and y variables
sns.pairplot(data=df,
        x_vars=["fatal_collisions_speeding", "fatal_collisions_alc"],
        y_vars=['premiums', 'insurance_losses'],
        kind='scatter',
        hue='Region',
        palette='husl')

plt.show()
plt.clf()
---------
# plot relationships between insurance_losses and premiums
sns.pairplot(data=df,
             vars=["insurance_losses", "premiums"],
             kind ='reg',
             palette='BrBG',
             diag_kind = 'kde',
             hue='Region')

plt.show()
plt.clf()


Building a JointGrid and jointplot
--------------------------------------------------
# Build a JointGrid comparing humidity and total_rentals
sns.set_style("whitegrid")
g = sns.JointGrid(x="hum",
            y="total_rentals",
            data=df,
            xlim=(0.1, 1.0)) 

g.plot(sns.regplot, sns.histplot)

plt.show()
plt.clf()
---------
# Create a jointplot similar to the JointGrid 
sns.jointplot(x="hum",
        y="total_rentals",
        kind='reg',
        data=df)

plt.show()
plt.clf()


Jointplots and regression
--------------------------------------------------
# Plot temp vs. total_rentals as a regression plot
sns.jointplot(x="temp",
         y="total_rentals",
         kind='reg',
         data=df,
         order=2,
         xlim=(0, 1))

plt.show()
plt.clf()
---------
# Plot a jointplot showing the residuals
sns.jointplot(x="temp",
        y="total_rentals",
        kind='resid',
        data=df,
        order=2)

plt.show()
plt.clf()


Complex jointplots
--------------------------------------------------
# Create a jointplot of temp vs. casual riders
# Include a kdeplot over the scatter plot
g = sns.jointplot(x="temp",
             y="casual",
             kind='scatter',
             data=df,
             marginal_kws=dict(bins=10))
g.plot_joint(sns.kdeplot)
    
plt.show()
plt.clf()
---------
# Replicate the above plot but only for registered riders
g = sns.jointplot(x="temp",
             y="registered",
             kind='scatter',
             data=df,
             marginal_kws=dict(bins=10))
g.plot_joint(sns.kdeplot)

plt.show()
plt.clf()