Difference between revisions of "Hiryuu Project Management"
Wtchua.2013 (talk | contribs) |
|||
Line 259: | Line 259: | ||
[[File:Hiryuu_Gantt_Chart.JPG|1000px|center]] | [[File:Hiryuu_Gantt_Chart.JPG|1000px|center]] | ||
+ | |||
+ | Midterm Changes: | ||
+ | * Exploratory Analysis extended by 2 weeks due to delay caused by data complications | ||
+ | * Clustering module shifted to start after recess week | ||
+ | * Time Series was brought forward and will end earlier than initial deadline (Wk 9 instead of Wk 11) | ||
+ | * Geospatial extended by another 2 weeks due to sponsor data/online data compatability issues | ||
+ | * Web app changed to Rshiny app. Reasons being that Rshiny is interactive, free, and can accomodate a wider range of analytics. Extend by 2 weeks because we are new to this |
Revision as of 17:01, 21 February 2017
|
|
|
|
|
|
Overview |
---|
Scope of Work
- Data Gathering and Scoping Data have been gathered from our sponsor which consists of data across 3 months and spanning across several countries. The data given to us are all in .csv formats.
- Research on Software and Proposal Preparation Our team has researched and discussed on the software and applications used in analysing and visualising the data with our supervisor and sponsor.
- Data Cleaning and restructuring
- Data Modelling
- Exploratory Analysis
- Geospatial Analysis
- Clustering
- Time Series Data Analysis
We will mainly utilise JMP Pro for the clustering, exploratory, and seasonality analysis and QGIS for our geospatial analysis. Our sponsor has expressed their preference for the final product to be in Power BI as they are familiar with it. In addition, they have expressed interest for the final product to be dynamic, meaning any input of future data can tap on the methods used to produce the results.
Currently we will focus on analysis on JMP Pro and QGIS and have plans to decide how to compromise the dynamic capabilities within our limitations.
After looking and familiarising ourselves with the data, we discovered the complexity of the data. The data is not only large in size, contains a wide variety, and also inconsistent across different countries. This inconsistency is due to the differences in language used for reporting as well as different systems used. As such, certain coding languages could not be used to process the values and other solutions have to be looked into.
Due to the complexity and variety of the data, we will be conducting data cleaning by removing duplicates, ensuring consistency by inserting the header column, merging the data files, filtering the data by countries and checking that the data is appropriate and ready for analysis.
Timeline
Tasks | Start | End | Teammates Involved | Status | ||
---|---|---|---|---|---|---|
Data Gathering and Scoping | ||||||
Gather Data | Week 0 | Week 0 | All | Completed | ||
Finalise Requirements with client | Week 0 | Week 0 | All | Completed | ||
Scope Project | Week 0 | Week 0 | All | Completed | ||
Research and Preparation | ||||||
Explore software | Week 0 | Week 0 | All | Completed | ||
Finalise Proposal | Week 0 | Week 0 | All | Completed | ||
Create and update wiki page | Week 0 | Week 0 | All | Completed | ||
Data Cleaning | ||||||
Data collection | Week 1 | Week 2 | All | Completed | ||
Data cleaning and restructuring | Week 1 | Week 2 | All | In Progress | ||
Resolve/remove incomplete data | Week 1 | Week 2 | All | In Progress | ||
Data Modelling | ||||||
Stage 1: Exploratory Analysis | Week 3 | Week 6 | All | Not Completed | ||
Stage 2: Clustering | Week 3 | Week 6 | All | Not Completed | ||
Stage 3: Time Series Analysis | Week 3 | Week 6 | Jouta | Not Completed | ||
Stage 4: Geospatial | Week 5 | Week 10 | Qianpin | Not Completed | ||
Interim Preparation | ||||||
Gather feedback from Client | Week 4 | Week 8 | All | Not Completed | ||
Prepare interim report and slides | Week 6 | Week 7 | Jouta | Not Completed | ||
Application Building | ||||||
Code the application | Week 8 | Week 11 | All | Not Completed | ||
Testing the application | Week 10 | Week 12 | All | Not Completed | ||
Gather feedback from Client | Week 12 | Week 16 | All | Not Completed | ||
Iteration | ||||||
Adjust analysis | Week 12 | Week 13 | Jouta | Not Completed | ||
Refine results to improve clarity | Week 12 | Week 13 | Wan Theng, Qianpin | Not Completed | ||
Final Preparation | ||||||
Prepare Research Paper | Week 14 | Week 14 | All | Not Completed | ||
Prepare Poster | Week 15 | Week 15 | Qianpin | Not Completed |
Work Plan
- Fortnightly meetings with Supervisor
- Monthly meetings with Sponsor
Midterm Changes:
- Exploratory Analysis extended by 2 weeks due to delay caused by data complications
- Clustering module shifted to start after recess week
- Time Series was brought forward and will end earlier than initial deadline (Wk 9 instead of Wk 11)
- Geospatial extended by another 2 weeks due to sponsor data/online data compatability issues
- Web app changed to Rshiny app. Reasons being that Rshiny is interactive, free, and can accomodate a wider range of analytics. Extend by 2 weeks because we are new to this