Category Archives: Data Visualization

Water scares city to Water Scarcity!

Title of this post is inspired from recent forward on social media about rain and water situation in Chennai. This forward also triggered the curiosity of data monks (a secret group) in Report Bee. Questions like?…

  • Where does Chennai get its water? What is its storage capacity?
  • How dependent is Chennai on Monsoons?
  • What will happen to water situation as Chennai’s population grows?
  • Is there a changing pattern in rains or weather?

We needed data to answer these questions. Luckily we had a good starting point for our data adventure – The website of Chennai Metro Water. This site has water levels of Chennai’s reservoirs (Lakes) for each day for the last 14 years (We wish it was 40 years of data!).

First step!

A Python script helped to retrieve the total volume of major reservoirs in Chennai. Once we had the data in the desired CSV format, it was play time 🙂

Where does Chennai get its water? What is its storage capacity?

We plotted various graphs to visualize the data and learn from it . The first graph, the total water levels from 2004 until today.


Observations from above graph, from 2004 up until 2012, the water levels have been fairly consistent across the year and these were years of water sufficiency. From 2013 onwards, there is a drastic fall in water levels across the 5 chief reservoirs which supply water to Chennai. This shortage of water did not stop even in 2015- December, the year infamous for Chennai floods.

Is there a changing pattern in rains or weather?

What we are seeing increasingly is a Monsoon that is holding itself back a little and then packs a punch too wet to handle in the months of Nov-Dec.

The truth is that when it rains so much in so little time, all water goes down the drain, literally. We need new ways to trap the monsoon efficiently.

Then, we visualized the 2004 to 2017 in an overlapping graph.

This visualization makes it is easy to compare the water levels across the years. In recent years, we run out of water faster and failed Monsoon threatens the city of drought. Validates the humour in the title of this post.

What will happen to water situation as Chennai’s population grows?

To understand this better, we calculated time taken for the water level to come to 3600 Mcft in a given year. Assuming that this level of water is a situation of concern. Listing the calculation for some of the years.

Jan 2006 – 8 months and 10 days to reach 3600 Mcft
Jan 2008 – 7 months and 20 days
Jan 2011 – 5 months to reach 5100 and then increases again
Dec 2015 – 8 months and 20 days to reach 3600 Mcft

In 2030, at full capacity in the 5 chief reservoirs, we’ll completely run of water in 9 months even if we get good rains. Chennai will face severe water scarcity in coming years.

  • Current Population of Chennai = 10,435,000, consumes 830 MLD/day
  • Predicted Population in 2030 = 13,921,000, will consume 1100 MLD/day
  • Reservoirs (Lakes) at Full Capacity = 300,000 MLD
  • Assuming no rainfall, No. of days the reservoirs will be sufficient for = 300,000/1,100 = 272 days (~9 months)

Do we need more desalination plants?  Do we need to build more reservoirs? Take a look at the table below.

Three were built in British era and other 2 much before! Are our catchment areas (reservoir locations) enough, do we need more?

How dependent is Chennai on Monsoons?

Monsoons are India’s lifeline. We used to be an agrarian economy dependent on rainfall for our survival in the global market. Today, years later, though the balance has shifted in favour of services more than manufacturing or agriculture, we are still dependent on these life-giving winds from the Indian Ocean. Even if not for agriculture in Chennai, it is to feed our rivers and put that glass of water on our dinner tables!

Big Data – How ‘Big’ is the data in Education?

Author: Manasa Raghuram

Big Data. Two simple words. When I first heard these words I wondered how much ‘Big’ data can education really have? And as I started working a year ago on Partnerships at Report Bee, I gradually realised that there is vast amount of educational data available for every student. But, was this being put to good use? After all, the data’s worth lies in the interpretation and the subsequent action, isn’t it?

School education, as we know, has evolved dramatically over the years. It is not just academics now. Schools offer many different programs ranging from life skills, leadership, robotics to yoga.  So, we have all this rich student data available with us: How do schools essentially make use of this?  With Report Bee in schools now, many different beautiful, insightful report cards are made possible at the click of a button!

Low income schools :

What happens with such vast education data in low income schools supported by impact organizations? More often students’ progress in those schools is never tracked since such data sets are unstructured, cluttered and lack insights. It is cumbersome to maintain and over a period of time this can lead to data loss when stored as hard copies or excel sheets.

Let’s consider a student named Priya. During Priya’s school life, she would have a unique data set related to her: general records with academics,daily attendance, health, co scholastic records, standardized test scores and so on . This in turn, becomes a huge data set for many students like Priya under an impact organization.

What if this huge unstructured, but extremely valuable education data is mined? Let’s for a moment, forget the analysis that can happen across the spectrum and focus on the students. We would be able to find simple patterns in student’s learning, allowing us to gain a deeper understanding of the individual academic performances and the factors influencing them which would in turn enable us to design more effective intervention programs.

What does Report Bee do?

The Partnerships team decided to actively foray into this new vertical to see how we can do what we do to help the many Impact Organizations in Education to make sense of the huge datasets they have.

So, What does Report Bee do for impact organizations?


Large Unstructured Education Data + Report Bee = Meaningful, Actionable Insights

Yes, we precisely do the above. The data given to Report Bee can be of any format, a text file, a comma separated file, a microsoft access database or any other format. Report Bee’s analytics engine is designed in a way to read these files of different formats and create insightful reports, that is visualized aesthetically for quicker and easy understanding.

Often times, different organizations have very specific objectives and have the need for a unique set of reports. That is when our data team steps in, providing customized visualizations with drilled down information regarding students’performance across subjects and across years. This can help organizations know the impact of their program on the students’ learning.

We know that the Best Decisions are Data Informed Decisions.

This information dissemination has in turn, helped the donors and other stakeholders with efficient capital allocation, eventually improving student learning.

Report Bee is all set to help create an impact in students’ learning by transforming how education data is read, understood, visualised and acted upon.

Big Data. Two small words with a big promise for education.

Boys perform well in Co-ed than single-sex schools

We belong to an era of data science, cognitive psychology and artistic visualization. At Report Bee, we performed an extensive analysis on single sex schools vis-a-vis co-ed schools and found that in Tamil Nadu, boys are performing better in co-ed schools than boys-only schools.

We took the last 4 years data (2012- 2015), and the findings showed us a trend.

In 2015, among the 6258 schools that appeared for the 12th board exams, the pass percentage of boys only schools is 59.8 whereas the pass percentage of girls only schools is 66.2. But in case of the co-ed schools, the pass percentage of boys and girls increases to 63.6 and 66.8 respectively.

In class 12, we found that boys in Co-ed schools on an average scoring 5.3% more than their counterparts in boys-only schools. Through a detailed analysis on each of the revenue districts, we observed a maximum average difference of 13.5% (~162 marks) in a Perambalur.

KANYAKUMARI, DINDIGUL, OOTY, NAMAKKAL, TRICHY, NAGAPATTINAM seem to have boys in boys-only school performing better than Co-ed schools, reason here is these districts are literacy districts and they perform well in any given condition.

When tested on girls in co-ed school and girls only school, there was an increase of 0.6%.It showed that girls are performing better than boys irrespective of being a single sex school or a Co-ed school.

Similarly In class 10, we found that boys in almost all the revenue districts in Tamil Nadu are performing better in Co-ed schools. Except Virudhunagar, all the other districts show a higher performance in average marks scored by the boys studying in co-ed school.

The performance of the students can vary from various reasons like peer pressure, discipline, employability, infrastructure, adolescence and so on.

Tamilnadu 12th Board – Spatial Analysis

12th Std boys12th std girls

Green colored districts are those in which the boys in Co-ed school have obtained a better average% than the boys-only schools. Red colored are the districts where the boys in boys-only schools have scored a higher average% than the Co-ed schools

Tamilnadu 10th Board – Spatial Analysis

10th std boys10th std girls