seabbs
P2P Blogger
Posts: 41
Likes: 22
|
Post by seabbs on Jul 31, 2017 9:50:03 GMT
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,707
Likes: 2,983
|
Post by michaelc on Aug 16, 2017 19:02:15 GMT
Hi
Nice tool.
I was trying to get a sense of how defaults vary by how many repayments were made.
I clicked summary on the loan book using variable to summarise as "defaulted" and variable to stratify by as "Repayments made". Doing that meant the x-axis showed repayments_made and y-axis seemed to be the total amount loaned to loans that had defaulted and made that number of repayments.
I wonder if there is a way just to show the number of loans that had defaulted for the given number of repayments_made? Or any other thoughts on how to show this kind of information?
Or even any thoughts from anyone on what I'm likely to expect? I thought I'd read somewhere to expect a bell curve with a low default rate very early on and very late on being higher in the middle?
|
|
seabbs
P2P Blogger
Posts: 41
Likes: 22
|
Post by seabbs on Aug 16, 2017 22:03:02 GMT
Hi Michaelc, Thanks - good to hear someone is using it! So first of all to answer your question. Setting should be: Variable to summarise = no. of loans, Variable to stratify by = repayments made, Variable to filter by = status, select = defaulted. It should look like this: www.dropbox.com/s/vb8d1e0ll45h3g3/Screenshot%202017-08-16%2022.54.43.png?dl=0This isn't the end of the story, if you don't adjust for time then you end up with a bias towards fewer payments (as there are many more newer loans). A simple adjustment for this is to limit the data to loans prior to Jan 2014. Link: www.dropbox.com/s/q2tvepwji89oyiv/Screenshot%202017-08-16%2022.55.20.png?dl=0There are other ways to do this obviously. If you want to investigate further you can facet by a variable, or filter by a second variable. For example, it might be interesting to see how this varies by risk band Hope that helps!
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,707
Likes: 2,983
|
Post by michaelc on Aug 17, 2017 19:38:00 GMT
Hi
Thanks again - very useful and interesting.
|
|
seabbs
P2P Blogger
Posts: 41
Likes: 22
|
Post by seabbs on Sept 27, 2017 10:37:33 GMT
|
|
seabbs
P2P Blogger
Posts: 41
Likes: 22
|
Post by seabbs on Oct 13, 2017 14:00:10 GMT
|
|
seabbs
P2P Blogger
Posts: 41
Likes: 22
|
Post by seabbs on Nov 5, 2017 17:10:43 GMT
Possibly my last update in a while as starting some part-time work at Funding Circle doing data analysis. I am freezing this account whilst I am there. FCdashboard will be staying up and as always a more up to date loan book can be uploaded by the user. Read how in the about section of the app (htpp://www.seabbs.co.uk/shiny/fcdashboard). Loanbook updated to the 05/11/17. Overview of loans originated by whole loan/part loan by loan amount: bit.ly/2hJ6pPL
|
|
|
Post by grahamreeds on Nov 5, 2017 22:14:39 GMT
Congratulations on the gig.
Hope you can finally give me an answer if they use automated testing!
|
|
markr
Member of DD Central
Posts: 766
Likes: 426
|
Post by markr on Nov 6, 2017 11:55:08 GMT
|
|
michaelc
Member of DD Central
Say No To T.D.S.
Posts: 5,707
Likes: 2,983
|
Post by michaelc on Nov 17, 2017 23:34:00 GMT
Possibly my last update in a while as starting some part-time work at Funding Circle doing data analysis. I am freezing this account whilst I am there. FCdashboard will be staying up and as always a more up to date loan book can be uploaded by the user. Read how in the about section of the app (htpp://www.seabbs.co.uk/shiny/fcdashboard). Loanbook updated to the 05/11/17. Overview of loans originated by whole loan/part loan by loan amount: bit.ly/2hJ6pPLGreat to hear that.
|
|