Methodology

       Vast Mini Challenge 2: Like a Duck to Water

Overview

Methodology

Observations

Conclusion

Back to Dropbox

Overview

Two datasets are provided to answer the questions for this challenge; one is “Boonsong Lekagul waterways readings” and other is “Chemical units of measure”. The former file includes information related to 106 chemical indicators of water quality and their readings collected from 10 locations across the preserve. There are total of 136,824 records of readings collected from January 1998 to December 2016. Belos is the snippet of dataset.

                          

Data Preparation

The above mentioned two data files are combined into one by matching measure name in both the files. Description of variables from the combined files is shown below.

                          

Data Cleaning

Initial data exploration of combined data files is performed in JMP. It is found that 9700 records of various chemical indicators have value of 0. Reading value of 0 is equivalent to missing value for any chemical measure. This can be problem if 0 is not removed from records and taken into tableau for visualization. Tableau treats 0 as value and plot these points in the visualization. To avoid that, records with value 0 is removed from dataset before loading into Tableau.

                          

Geo – Location Mapping

In addition to two data files, an image file “Waterways Final” with location of various water sensors across the preserve is also provided. This image file is added to tableau workspace as a background image to extract the coordinates of every location of sensors. New excel file is created as “Location.csv” with column names as location, X and Y. The point that represents each location of sensors in the map is annotated to get X and Y coordinates. The values of coordinates are then manually included into the excel file.

                          

Tools Used

• Microsoft Excel
• JMP Pro
• Tableau