SpendAdvisor is a simple application designed to answer one question: Can I afford to spend $x right now? It uses the Geezeo API and thus requires access thereto.
Before running the application, you will need to set up a few
things. (Note: This is a Rails 4.1.0 app. Ostensibly, the only
change required to be Rails 3 compatible is to swap out the use of
secrets.yml.)
In order to not leak the following sensitive information, the
secrets.yml file has been used to grab some values from the
environment, like so:
gzo_api_key: <%= ENV["GZO_API_KEY"] %>
gzo_cust_id: <%= ENV["GZO_CUST_ID"] %>
gzo_domain: <%= ENV["GZO_DOMAIN"] %>
gzo_user_id: <%= ENV["GZO_USER_ID"] %>
So in your .bashrc or similar, export these four keys using the
appropriate values provided by Geezeo, for example:
export GZO_API_KEY=pneumonoultramicroscopicsilicovolcanoconiosissupercalifragilisticexpialidocious
export GZO_CUST_ID=example
export GZO_DOMAIN=dev.example.com
export GZO_USER_ID=auntmaeMake sure you source your startup file before running tests or the rails server.
The Geezeo user will have to have a checking account that is reporting a balance, as well as Cashflow Bills and Cashflow Incomes.
For the latter, a rake script has been created to generate a few sample cashflow bills and incomes. However, the accout balance cannot be frobbed as far as I can tell. (At least not from the API, for obvious reasons.) To run the script:
$ rake data:all
Two rake tasks will be run: one for deleting all cashflow bills and incomes, and one that recreates them.
You may also create your own cashflow data by modifying the
lib/tasks/test_cashflow_bills.yml and
lib/tasks/test_cashflow_incomes.yml files accordingly, and running
the rake script.
To run tests, simply run rake test test/models/gzo_test.rb. There
are no tests for the controller or UI yet.
Also, make sure to re-run the rake data:all script after running
tests to reset the generated cashflows as per the seed data (YAML
files).
$ rails server should do the trick! This is what it is supposed to look like:
These are features I had in mind but have not gotten around to implement.
Rather than a Yes/No answer, I thought it would be a good idea to have four different answers, such as:
- Definitely not
- Probably not
- Yes, but barely
- Yes, all things being equal
Or something to that effect. Also, each answer would be accompanied by an appropriate animated reaction gif. :)
Currently, the minimum balance that needs to be maintained on the account is set to $100. Anything below that will cause a "No" answer. The user should have the option to set their own amount.
Since cashflow forecasting is trick at the personal level, there should be more wiggle room for unforeseen circumstances. The user should have the option to set a "risk multiplier" which would decrease the confidence level in the answer given. This could also be used by users with irregular income to adjust for gaps and such.
In reality, the cashflow forecasting should should consider historical data (actual cashflow events) in order to validate the reality of a cashflow based decision. If your cashflow for the past 6-12 months has been completely dissonant with the planned cashflow, the decision should reflect this or at least inform the dissonance.
Alternatively, the cashflow could simply be generated or seeded based on historical data, and the software would continue to work as is. However, this requires that the user has enough transactions on which to perform statistical analysis.
Show the cashflow bill and week on which the account balance would go below minimum account balance.
Also potentially show which cashflow income would cause the account balance to go back above the min-acct-bal, and how long that would take.
A nice feature would be to have a nice visualization of your cashflow forecast, and how and immediate expense would affect it. (Alternatively, simply create a new cashflow with a 'once' frequency and re-run the calculations.)