Interactive online dashboard for survey results
Background : The client is a health NGO working to fight risks of non-communicable diseases (in particular, diabetes) in rural India. Their schools program conducts programs in schools across Karnataka and Andhra Pradesh to make behavioural impact related to eating habits and level of physical activity. The program impact is measured by conducting surveys.
Requirement : The client wanted a way to easily sift through the data, understand the high level messages and drill down to details as required.
Solution : In this project, I built an online dashboard into which multiple survey datasets can be uploaded and the survey data would be displayed in an organized and intuitive way. The user can choose the data set to be viewed and the results are shown grouped in relevant categories and across timelines.
This dashboard helped the users quickly understand what the survey results were showing and drill down into details for further analysis. The charts required for reporting can be easily taken from this dashboard.
Snapshot : The chart below is purely illustrative and do not reflect any client data.
Visualization for Healthy Workplace project reporting
Background : The client conducts research on workplaces with respect to its employee health and wellness programs. Every year nominees are evaluated based on a data centric framework and Platinum awards are presented to the winners.
Requirement : Participant company data needed to be analysed and presented for the evaluation of Platinum awardees based on the assessment framework.
Solution : The project involved organizing the database to manage the company datasets and creating a tool to generate the visualization charts needed to present the analysis.
Predictive modelling for commuting time series
Background : In 2017, Uber introduced Uber Movement – a site where you can find time estimates on how long it takes to get from one place to another. These estimates are taken from aggregated data of Uber trips and are availale across different times of the day, days of the week and months of the year. In 2018, the Movement tool was rolled out for Delhi, Mumbai and Bangalore.
Project : The objective was to evaluate ways in which city planners can use this database to better plan activities such as infrastructure development and maintenance and also do a cost-benefit analysis. A predictive model was used to project the daily average commute time between two points in Bangalore.
Snapshot : Below, in the first chart is the data extract from site for the time taken between two points. And the second is the prediction analysis of this time series using the Prophet model.
Interactive dashboard for world electricity sources
Project : This dashboard was built to present and drill down into the sources of electricity production across countries and geographic zones. The objective was to understand how R Shiny and ggplot libraries can be used to build interactive dashboards.