Skip to content

Python Project which was aimed at data wrangling, analyzing and creating visualizations using python in Jupyter notbook .

Notifications You must be signed in to change notification settings

drvalerioh/DataVisulaizationpython

Repository files navigation

DataVisulaizationpython

Prosper Loan Data

by Valerian Lunale Makanga

Dataset

This dataset contains attributes of APRs and variables of 113937 loans and hence attributes included original loan amounts, borrower's Prosper rating, loan term, borrower's stated monthly income whis are extensively explored in the dataset analysis.

This dataset contains number of loans in loans amounts, borrower rate /interest rate, current loan status, and borrower income.

From the loans borrowerd, all are determined from different ratings ,terms of paymenst and status of specific LoanAmounts.

However, their different listings which are associated with the loans for qualified applicants and categories.

Summary of Findings

Univariate explorartions1. Explanations:

in the prosper dataset, the current loans status are the most wherethe completed loans are the second largest section in Prosper however,

all the PastDue loans are the least in the dataset and they are listed according to their days past due with (1-15 days) being the most.

the cancelled loan status have a nill/0 value count.chargedOff loans and Defaulted loans are still a high value of count in the LoansData.

Observation.

In general , PropserData loans are completed than at higher rate as compared to Defaulted status. The StatedMonthlyIncome, the histogram gets clustered to the left due to uneven distribution of data.

the Employed are the most borrowers in the prosperLoanData then the Full-time borrowers whereas other categories are small parts of borrowers. most of the loans borrowerd from prosperLoanData set are Current Loans and in Term variable most of loans took 36 months as the longest period of borrowers.

Bivariate Exploration.

In this section, investigate relationships between pairs of variables in your data. Make sure the variables that you cover here have been introduced in some fashion in the previous section (univariate exploration). In the listing Category, Debt Consolidation, Home Improvement, Business, Auto and Other have less loans borrowed. the Auto listings category has the list borrowed numberand Debt Consolidation the highest number in the listing. Explanations and Observations: in the current the most ProsperRating is C and according to the findings while HR is the list at the time the listing created. in Completed Rating, D is the highest ProsperRating since the loan listing was created. the A ProsperRatings appear relatively higher for the Completed and the Current. In the Loanstatus, the current borrowed loans have the highest ProsperRating in all the other ProsperLoanData set.

Explanation and Observations.

In the Loanstatus Listings: the Current Loanstatus is the highest in the Graphs of Borrowers Completed and Completed Loan status. Debt Consolidation Graph is the Highest in Current Loans and relatiely low in Completed and Defaulted Loans in the prosperLoanData set.

Observations and Explanantions.

From the Graph plotting visual, its noticed that the Current Loanstatus in the listings are the largest scores in the dataset as compared to Defaulted and Completed LoansAmounts.

Observations and Explanations.

from the dataset Graph plotting, the Borrowers with the status Retired, Part-time, Not-Employed and others are lower in loans Ratings as compared to Self employed. the Borrowers with the status of Employed have a higher rating of Loans from prosperLoanData company across all the other Ratings and listings since created in July 2009.

Explanations Key Findings are:

In the prosperLoanData set, when i plot LoanStatus, its evident that Current Loan Borrowers are the majority in the Listings since July 2009, and relatively Loans that are Completed are equally higher as compared to Defaulted Loans.

in Employment Status against Loans borrrowerd, its noticed that most borrowers who's status is Employed tend to have the highest propability of acquiring Loans with prosper loaning company, consiquently when Not -Employed, Retired or Part-Time you have little chances of borrowing Loans from propser company.

Its evident that ,for a good and higher Loans Rating ,you need Employment status hence their is a co-relation between Loans status and Term used in paying which is 36 months for most borrowers.

from the dataset Graph plotting, the Borrowers with the status Retired, Part-time, Not-Employed and others are lower in loans Ratings as compared to Self employed.

the Borrowers with the status of Employed have a higher rating of Loans from prosperLoanData company across all the other Ratings and listings since created in July 2009

Multivariate Explorations:

Explanation and Observation:

from prosperLoanData plotting, LoanOrdinalAmount is the Origination amount of the Loan.

in the dataset, most loans that are Defaulted have relatively high LoanOrdinalAmount than Completed but lower rating than current Loans in the dataset.

Current loans will still be the highest amounts of loans in prosper company, similarly its noticed that HR loan Ordinal ratings are the list ratings for both the Current, Defaulted and Completed loans in the dataset prosper company which is different case for E PropserRatings that Curent loans appeare to be the list hence list that E as the least prefered Rating by any of the Loans status by borrowers.

From further explorations:

In LoanOrdinalAmounts, Auto Listing Category in prosper Company tend to have a simillar trend in borrowing of Loans, besides the loan credits Business and Home Improvements have a similar size in Loans borrowing amounts.

Debt Consolidation Category in prosper company is similar and of lower ratings is completed and Defaulted Loans as compared to Curent Loans which have a higher Listing Amounts across all the Other, HomeImprovement, Auto and Business ratings.

Consiquently, all the ListingCategories are above the quartely threshold of the highest rating whish are the Current Loans. StatedMonthlyIncome is the income the Borrower stated at the time the listing was created.

Similarly the loan Term of 36 months is still dorminant in the dataset of loan Borrowers, hence with propser rating the three Loan Terms of 12, 36 and 60 in months which co-relates with the amounts of loans significantly.

However, the Term and the effect do not co-relate in the borrowing of loans and monthly terms associated.

From Observeable relationships.

There are no exisiting relationships between the Term of loans and Rating rather the Status which is either , Employed, Self-Employed,Part-time among others that determine.

in the analysis, with beteer ratings the loan amounts of the Terms increases as the terms and loan amounts also increases. In the prosper rating its noticed that , from HR listings, in StatedMonthlyIncome and LoanOrdinalRating, 60 months loans have a higher Ordinal rating as compared to 36 months for LoanAmountsIncome dataset, Therefore, the multivariate explorartion reveals that In Loan duartion of 36 months ismost prefered for the Current loans which are also the highest in the borrowing of loans.

positivelt i can relate the term on 36 months to Employed stattus borrowers who are also the more prefreable in the prosper company since they also are the majority of the loans borrowers in the dataset.

Key Insights for Presentation

From the key presentations of exploratrions, i focused my exploration and findings on Loan Status and relevant defining factors that lead to loans being issued to borrowers. Additionally, the APR ratings and terms of listings of loans are related to the extend at which of awarding loans and listings.

I also did investigate the effect of the loan ratings between the loan amount as thew effects of the ratings on the relationships relevant to the term of loan allocation from the prosper. From prosperloandata, there binding factors that determine the loan borrowing and the ratings that determine the allocations interms of months in loans.

However, all the loans from prosper company depend on the loanstatus of Employed since its evident from the explorations that all borrowers from that category are the most preferd for loans in the company.

most loans in the company are also current loans borrowerd and defaulated loans being the least in the loaning from the company.

About

Python Project which was aimed at data wrangling, analyzing and creating visualizations using python in Jupyter notbook .

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published