gps-taxi: A repository from embruze

#The Master Dataset Download the Master Dataset 17-18 GB .csv file

Download the July subset of the Master Dataset ~1.5 GB

##Latest Maps ###Taxi 2013 Distance from Roadbed by Census Block 2010 & Taxi 2013 Count Taxi Pickup and Dropoffs by Census Block 2010

#Working Maps and Code ##Maps

##Concept ###Google Earth screenshot ###Building and neighborhood geography ###Building Volume Spread over the entire neighborhood

##Plots

##Input Spatial Data ###Roadbed Data

###Census Blocks 2010 - NYC DCP

####Get geoid code

def boroCountyFIPS(boro,tract,block):
  st = '36'
  if boro == 1:
    return st + '061' + tract + block
  elif boro == 2:
    return st + '005' + tract + block
  elif boro == 3:
    return st + '047' + tract + block
  elif boro == 4:
    return st + '081' + tract + block
  elif boro == 5:
    return st + '085' + tract + block
  else:
    return 'X' + tract + block

##Code and Repo's

#Completed Maps and Code

#Data ###Sample January 1, 2013 Pickups Only

###Richie's Data

#To Do:

###DS Here’s the next steps I’m doing: New vars:

Create Census Block flag variable for Park or No Park
Create Census Block flag variable for Buildings or no Buildings.
Create Census Block Size variable. (Maybe also measure of possible maximum distance across block, or measure of block compactness)

New Maps:

Create Median Distance to Roadbed groupby dataset.

Database:

Coordinate with Amazon about possible Free account, setup PostgreSQL (SQLite or equiv. ) database.
Work with Sankara to get SQL queries to pull data out of database.
Alternatively is this doesn’t work we can look into https://code.google.com/p/gcsvsql/

Other Group tasks:

(RC, EB) Formulate questions for Prof. about analytical methods (regression vs. other methods).
(SG) Work on SQL queries for random selections (contingent on either database or SQL/csv wrapper).
(All or someone could volunteer to take lead getting started) Start the Presentation outline (https://github.com/stat4701-edav-gps/presentation) - can be edited inline via GitHub website. Edit README.md file.
(All or someone could volunteer to take lead on scheduling) Figure out which day/time/place for weekend of April 25-26.

Steve suggestion - log transform

####Complete Test sampling. 1 month. July.

EB, RC, SG, DS

Sankara, Emilie and I met yesterday and here are next steps. Is this possible to be done over the weekend?:

Richie - Pull one day's worth of data and try to do some clustering. Let's use what Danny posted for Feb 26, 2013
Danny - Use that data to plot on a map. Danny, did you have ideas on what specifically will be useful? We thought we'd like to see how the plot would look if overlaid and how popular some places are.
Emilie/Sankara - Write up the plan.

Let me know if I missed anything. Let's reconvene Tuesday.

#Outline:

Front Matter

Title page (page 1)
Executive summary (page 1)
Table of contents (page 2)
Lists of figures and tables (page 2)

Introduction

Statement of the research problem
Project purpose, background, and context
Main objectives and contributions of the study
Spatial questions and hypotheses
Study area description, location, and context

Research Design

Literature Review (at least 3 scholarly)
Conceptual framework and definition of key terms
Description of concepts

Methods

Description and operationalization of data
Data analysis and modeling (bulleted)

Results & Discussion

Results of data analyses
Presentation of results in text, figures, and tables
Review of results in relation to spatial hypotheses and research problem
Discussion of results in the context of the project background
Conclusions and recommendations based on results
Recommendations for future studies

References

Data Source References (last page)
Works Cited
Appendices (if neccessary)

embruze/gps-taxi