Difference between revisions of "IS428 AY2019-20T1 Assign Tan Sok Yi DataTransformation"
Line 52: | Line 52: | ||
[[File:coordinates-3.png|600px|center]] | [[File:coordinates-3.png|600px|center]] | ||
<br> | <br> | ||
− | * c. Click on View Data under Analysis and the data can be exported as a | + | * c. Click on View Data under Analysis and the data can be exported as a CSV file. |
[[File:coordinates-4.png|600px|center]] | [[File:coordinates-4.png|600px|center]] | ||
+ | |} | ||
+ | |||
+ | === 2. Transposing of the data === | ||
+ | {| class="wikitable" | ||
+ | |- | ||
+ | ! Problem 1 || Transposing of the data | ||
+ | |- | ||
+ | | Issue || The dataset provides multiple responses collected from different users at the different timestamp. However, based on the structure of the data, it is difficult to analyse the data and apply filters on the visualisation based on the different damage area. | ||
+ | [[File: transpose-1.png|600px|center]] | ||
+ | |- | ||
+ | | Solution || To solve this issue, we will make use of Tableau Prep to transpose the different columns (buildings, medical, power, roads and bridges, sewer and water and shake intensity) into one column for better analysis. | ||
+ | <br> | ||
+ | * a. Create a new flow by dragging the dataset into Tableau Prep. | ||
+ | |||
+ | [[File:transpose-2.png|600px|center]] | ||
+ | <br> | ||
+ | * b. Add a pivot and drag the fields (buildings, medical, power, roads and bridges, sewer and water and shake intensity) to swap the different columns into different rows. | ||
+ | |||
+ | [[File:transpose-3.png|600px|center]] | ||
+ | <br> | ||
+ | * c. Add output and the transposed table will be shown. After pivoting the table, the different area is now under one column named “Damage Area” and its respective value is under another column “Damage Value”. Save the file as a CSV file. Then, we will be able to create the respective visualization necessary for the assignment. | ||
+ | |||
+ | [[File:transpose-4.png|600px|center]] | ||
+ | |} |
Revision as of 21:28, 9 October 2019
IS428 Visual Detective: Crowdsourcing for Situational Awareness
|
|
|
|
|
Contents
Data analysis and cleaning process
Before creating the visualisation, analysis of the dataset is done to understand the different variables and attributes. One dataset, containing reports of shake intensity and the intensity of the damage in the different areas (buildings, medical, power, road and bridges, sewer and water) in the various neighbourhood, in an interval of 5 minutes. Shake maps were also provided which allowed a better understanding of the coverage of the earthquake.
Geographical data is available in MC2 where a shp file showing the coordinates of the different neighbourhood. This dataset would be useful in creating the visualisation for MC1. For the datasets given, it needs to be clean and process first to prepare the data for the visualization.
Data Cleaning
1. Getting the coordinates of the different locations
Problem 1 | Getting the coordinates of the different locations |
---|---|
Issue | Geographical data is needed to create a map visualisation which will be useful in analysing the dataset. However, the exact location of the neighbourhood cannot be plotted as the coordinates of the different locations are not available in the dataset. |
Solution | To solve this problem, the data is available as a shp file in MC2 can be used to retrieve the coordinates for the different location. After retrieving the coordinates, a polygon on a map can be created.
|
2. Transposing of the data
Problem 1 | Transposing of the data |
---|---|
Issue | The dataset provides multiple responses collected from different users at the different timestamp. However, based on the structure of the data, it is difficult to analyse the data and apply filters on the visualisation based on the different damage area. |
Solution | To solve this issue, we will make use of Tableau Prep to transpose the different columns (buildings, medical, power, roads and bridges, sewer and water and shake intensity) into one column for better analysis.
|