Python-predictive-model

using google predictive model API and output prediction with python script
http://sunlit-pipe-95822.appspot.com/
deployed with google appcfg to appspot
todos future - front end view, sample data view, load file system for new queries

Traning Docs

---

-sample training on language prediction api

https://cloud.google.com/prediction/docs/hello_world

-try out the cloud predictive spreadsheet

https://docs.google.com/spreadsheet/ccc?key=0AisEGDqMba8MdFlMSmcyaXBfT1dUN1dPLTZXNUd3WUE#gid=7

-try out the hosted models examples

projectnumber:414649711441
model:sample.languageid
https://developers.google.com/apis-explorer/?hl=en_US#p/prediction/v1.6/prediction.hostedmodels.predict

-clone and modified to work with clientid/secret

https://github.com/GoogleCloudPlatform/prediction-try-java-python

figured out flow in OAuth2.0 to allow request access
once have access stored in backend, then you can make requests

my python script: main.py

---

usage:

script: use the ./run.py to skip the manual step

manual: while inside the folder and hav google_appengine at home directory, in terminal run
~/google_appengine/dev_appserver.py ./
open browser at http://localhost:8080

it would ask to auth access for OAuth2 for the first time

click accpet and it would query and display results

to add/modify the query prediction, change the query.csv

file description:

run.py - script to start server and browser

main.py - codes for server and query

data_io.py - codes for processing csv query

access.py - variable for accessing models in google prediction api

dataset sat scores- https://catalog.data.gov/dataset/sat-results-e88d7

functional design:

routes setup at the bottom

on default /, directs to MainPage

if credential doesnt exist, opens up flow in OAuth and redirects to AuthHandler
AuthHandler checks if credential exists, if so returns to main

if credential already done, then process those the query for each row in query.csv