kshirsagarsiddharth

def univariate_hist_with_hue(x = 'person_age', hue = 'person_income'):
    plt.figure(dpi = 130)
    sns.set_style('whitegrid')
    return sns.kdeplot(x = x, data = df, fill=True, palette = 'crest', hue = hue).set_title('Univariate visualization of Quantative Variable with hue')

D = interact(univariate_hist_with_hue,
             x = widgets.Dropdown(
                 options = ['person_age','person_income','person_emp_length','loan_amnt','loan_int_rate','loan_percent_income','cb_person_cred_hist_length']
                 ),
             hue = widgets.Dropdown(

kshirsagarsiddharth

def univariate_hist(x = 'person_age'):
    plt.figure(dpi = 130)
    sns.set_style('whitegrid')
    return sns.kdeplot(x = x, data = df, fill=True, palette = 'crest').set_title('Univariate visualization of Quantative Variable')

C = interact(univariate_hist,
             x = widgets.Dropdown(
                 options = ['person_age','person_income','person_emp_length','loan_amnt','loan_int_rate','loan_percent_income','cb_person_cred_hist_length']
                 )
            )

kshirsagarsiddharth

def scatter_plot_int_with_hue(x = 'person_age',y = 'person_income', hue = 'loan_grade'):
    plt.figure(dpi = 120)
    sns.set_style('whitegrid')
    return sns.scatterplot(data = df, x = x,y = y, alpha = 0.6, hue = hue, cmap = 'Set2').set_title('Visualize Relation Between 2 Quantative Variables with Hue')

B = interact(scatter_plot_int_with_hue,
             x = widgets.Dropdown(
                 options = ['person_age','person_income','person_emp_length','loan_amnt','loan_int_rate','loan_percent_income','cb_person_cred_hist_length']
                 ),
             y = widgets.Dropdown(

kshirsagarsiddharth

def scatter_plot_int(x = 'person_age',y = 'person_income'):
    plt.figure(dpi = 120)
    sns.set_style('whitegrid')
    return sns.scatterplot(data = df, x = x,y = y, alpha = 0.6, ).set_title('Visualize Relation Between 2 Quantative Variables')
A = interact(scatter_plot_int,
             x = widgets.Dropdown(
                 options = ['person_age','person_income','person_emp_length','loan_amnt','loan_int_rate','loan_percent_income','cb_person_cred_hist_length']
                 ),
             y = widgets.Dropdown(
                 options = ['person_age','person_income','person_emp_length','loan_amnt','loan_int_rate','loan_percent_income','cb_person_cred_hist_length']

kshirsagarsiddharth

## Highliting losses in the dataframe 
def color_negative_values(value): 
    """
    This function takes in values of dataframe 
    if particular value is negative it is colored as redwhich implies loss
    if value is greater than one it implies higher profit
    """
    if value < 0:
        color = '#ff8a8a'
    elif value > 1:

kshirsagarsiddharth

_ = interact(plot_histogram,
             palette = widgets.Dropdown(
                 options = ['pastel','husl','Set2','flare','crest','magma','icefire']
                 ),
             kde = widgets.RadioButtons(
                     options = [True,False],
                     disabled = False),
            hue = widgets.ToggleButtons(
                options = ['categories','other categories'],
                tooltip = ['categories','other categories'],

kshirsagarsiddharth

def plot_histogram(bins = 10, hue = 'categories', kde = False, palette = 'Blues', x_range_1 = (-3,3)): 
    """plots histogram
    params:
    =======
    bins: int
        histogram bins
    hue: str
        categorical columns to color 
    kde: bool 
        wether to show kde plot 

kshirsagarsiddharth

als = ALS(maxIter = 10 ,
          userCol = "user_id", 
          itemCol = "isbn_indexed", 
          ratingCol = "book_rating",
          nonnegative = True, 
          coldStartStrategy = 'drop')

from pyspark.ml.tuning import ParamGridBuilder,CrossValidator 
grid = ParamGridBuilder().addGrid(als.rank, [10,30])\
                               .addGrid(als.regParam, [0.2,0.01,1,2])\

kshirsagarsiddharth

from pyspark.ml.recommendation import ALS 
from pyspark.ml.evaluation import RegressionEvaluator 

training, test = indexed.randomSplit([0.8,0.1])

als = ALS(maxIter = 10 , 
          regParam = 0.9, 
          userCol = "user_id", 
          itemCol = "isbn_indexed", 
          ratingCol = "book_rating",

kshirsagarsiddharth

from pyspark.ml.recommendation import ALS 
from pyspark.ml.evaluation import RegressionEvaluator 
from pyspark.ml.feature import StringIndexer 

indexer = StringIndexer(inputCol = "isbn", outputCol = "isbn_indexed") 

indexed = indexer.fit(filtered_with_location).transform(filtered_with_location)\
                 .withColumn('isbn_indexed',F.col('isbn_indexed')\
                 .cast("int"))\
                 .drop('isbn')