Skip to content
This repository has been archived by the owner on Oct 6, 2021. It is now read-only.

Latest commit

 

History

History
63 lines (53 loc) · 3.73 KB

dataDictionary.md

File metadata and controls

63 lines (53 loc) · 3.73 KB
Raw

Data Dictionary

Data Files available:

  • Raw data: coc.csv
  • Un-duplicated "Problems" file : problemCategoryDictionary.csv
  • Complaints rolled up to day: "rolledComplaints.csv"

Raw Data: coc.csv

This is the raw data.

Details

Name Description Type Missing Value
Serial ID (Auto-Increment ID) Numeric -
Complaint.Type Type of the Complaint. 93 possibilities String, Categorical -
Area Denotes the area of the complaint. 260 possibilities String, Categorical "NA"
Locality Location at the "Locality" level. One level below "Area". 3392 possibilities (currently, may change in the future) String, Categorical "nan"
Street Location at the street level. One level below "Locality". ~25k possibilities String, Categorical "NA"
Region The region of the complaint. This is automatically filled using user entered details. Possible Values: ["Central Region", "South Region", "North Region"] String, Categorical -
Zone The zone of the of complaint, like "Region", but at a lower level. Possible values: ['N01' 'N02' 'N03' 'N04' 'N05' 'N06','N07' 'N08' 'N09' 'N10' 'N11' 'N12', 'N13' 'N14' 'N15'] String, Categorical -
Ward The ward of the complaint. Like "Zone" but at a lower level. There are 200 wards. String, Categorical -
Complaint.Details Text describing the complaint. Not always filled by the user, so the data contains a lot of duplicates. String, Text -
Email Anonymized Email String, Categorical "00e3902431681886db4cf60de2db9188" is no / empty email
Mobile Anonymized Mobile Numbers String, Categorical -
Department The department the complaint is assigned to. 48 possibilities. String, Categorical -
Complaint.Date The date of filing the complaint. dateFormat = "%m/%d/%Y %H:%M:%S" (python) DateTime "NA"
Resolution.Date The day of complaint resolution. dateFormat = "%m/%d/%Y %H:%M:%S" (python) DateTime "NA"
Number.of.days.for.resolution Derived attribute. Number of days taken to resolve complain. Set only if Is.Resolved = "YES" Numerical "NA"
Range Derived from Number.of.days.for.resolution. Set onlu if Is.Resolved="YES". Possible Values: '(16-30 days)', '(3-7 days)', '(8-15 days)', '(<3 days)','(>30 days)' String, Categorical "NA"
Is.Resolved Whether it is resolved "YES" or not "NO" String, Categorical -

Un-duplicated "Problems" file : problemCategoryDictionary.csv

This is generated from the raw data in the following way:

  • Exact Duplicates in the columns ['Complaint.Type', 'Area', 'Locality', 'Street', 'Region', 'Zone', 'Ward', 'Complaint.Details', u'Email', u'Mobile', u'Complaint.Month', "Complaint.Year", u'Complaint.Day'] are removed
  • Only resolved complaints are selected
  • Exact Duplicates are removed in these columns ["Complaint.Type","Street", "Resolution.Year", "Resolution.Month", "Resolution.Day"]

It has the same structure as the raw data, except for additional derived data. They are listed below:

  • Fields derived from "Complaint.Date":
    • 'Complaint.Year'
    • 'Complaint.Month'
    • 'Complaint.Day'
    • 'Complaint.Hour'
  • Fields derived from "Resolution.Date":
    • 'Resolution.Year'
    • 'Resolution.Month'
    • 'Resolution.Day'
    • 'Resolution.Hour

Complaints rolled up to day: "rolledComplaints.csv"

This is aggregated over the complaints, and rolled up so that all complaints are summed up according the the type of the complaint, on a per day basis. The columns are:

  • "Complaint.Year" - Explained Earlier
  • "Complaint.Month" - Explained Earlier
  • "Complaint.Day" - Explained Earlier
  • "Complaint.Type" - Explained Earlier
  • "Ward" - Explained Earlier
  • "Department" - Explained Earlier
  • "Count" - The number of complaints