The mighty statistician CR Rao has passed: Rao has conquered all
Corinthians 15:55 - 57 asks thus, “O death, where is thy sting? O grave, where is thy victory?”.
A remarkable man of Indian nationality by the name of Calyampudi Radhakrishna Rao retired from the Indian Statistical Institute after serving there for 40 years. He moved to the US in 1979 where he served the statistical community for another 44 years. His career spanned 84 years in the field making an incredible contribution to statistical theory and practice.
On the 22nd August the world and the statistics community was robbed of Rao, at the age of 102. Fondly known as CR he was a member of faculty at Buffalo University until death cut the visible to the eye ribbon of his career.
Yet death can only resuscitate to the living the fond memories of how CR Rao fulfilled his mission on earth, benefiting nations and the statistical community over his illustrious career of 84 years.
It is this context of death that vividly brings to me the memories of the lanky figure dressed in Indian attire and delivering a powerful acceptance speech of his mastery of the subject in an accessible fashion. In statistical speak, this was exactly almost two decades ago.
The University of Pretoria conferred an honorary doctorate to Professor Rao in 2004. I had the pleasure of hosting him at Statistics South Africa as the then Statistician-General while he was here in South Africa to receive the doctorate.
In his seminal reception speech at the University of Pretoria he reminded the audience of the human spirit of president Nelson Mandela and his love for statistics. He said he was surprised by the rare advice of a politician to society when he had a meeting with him. Madiba’s counsel was study statistics, practice it consistently by doing it a little every day.
CR had cut his teeth and rubbed shoulders at the Indian Statistics Institute with some of the greatest of minds. One of them was the towering figure not only in India, but in the whole world. This was none other than Prasanta Chandra Mahalanobis after whom the International Statistics Institute prize from the Indian government is named.
Professor Rao, in his presentation, chuckled at the error of specification, which even the best can make resulting in incredibly ridiculous results. He emphasised that model specification is a crucial stage in study design and understanding the context is even more crucial. Yet even the greatest Mahalanobis fell prey to this omission.
One of the studies by Mahalanobis, on food consumption patterns in India, revealed entertaining results, which were of women eating negative rice. This was an embarrassing finding for Mahalanobis. Upon closer look of the results, Rao said the error was caused by the absence of context that should have been embedded in the model design.
When there are shortages of rice, the mother would rather not eat and allow her share of rice to be eaten by the husband and the children. Such a consideration would have constrained the model to have the y-intercept to be non-negative.
Mahalanobis, having not taken into account this known observation amongst mothers in India, a matter which he had to have known better as he immersed himself in what has become the longest statistical time-series in the world, was embarrassed by this surprising result.
CR chuckled about this like a naughty school boy. The study was not published.
An equivalent study was undertaken elsewhere in the world that tried to model driving shares between spouses. The result was even more bizarre. The equivalent of the negative rice was driving in reverse for the wife and driving forward for the husband. The fact of the matter here again failed to take into consideration that the wife did not drive. For that reason, the model should have set the y-intercept as a non-negative number.
It is said, “On the basis of Rao's recommendation, the Asian Statistical Institute (ASI), now known as the Statistical Institute for Asia and Pacific, was established in Tokyo to provide training to statisticians working in government and industrial organisations”. Many statisticians in government especially have gone through that institute.
A tribute to Rao when he reached 101 years reads like a book from the future life of virtual reality. By the age of 25 CR Rao had accomplished a lot.
Bradley Efron, a Professor Emeritus of Statistics, further writes that “the 25-year-old Rao in 1945 introduced differential geometry into statistical inference, opening up the burgeoning field now called information geometry.”
Rao distances combined with conformal mappings are seen in modern applications such as virtual tourism.
Shun-ichi Amari, a Japanese mathematician at Tokyo University, writes, “Rao’s initiation of information geometry is one of the many achievements for which he was awarded the US National Medal of Science. Information geometry has grown to become an important tool not only in statistics, but also in artificial intelligence, data science, signal processing, physics, and many other fields since it elucidates the fundamental structure of the manifold of probabilities.”
“O death, where is thy sting? O grave, where is thy victory?” Rao has conquered all.
Dr Pali Lehohla is the director of the Economic Modelling Academy, a Professor of Practice at the University of Johannesburg, a Research Associate at Oxford University, a board member of Institute for Economic Justice at Wits and a distinguished Alumni of the University of Ghana. He is the former Statistician-General of South Africa.
BUSINESS REPORT