Adata scientist makes his argument for using trees and random forests: “trees and random forests are superior to machine learning methods, such as support vector machines and neural networks, because trees are unaffected by monotonic transformations of the explanatory variables.” how would you go about testing the transformation-invariance argument (with python and scikit-learn, for example) ?

Explanation:

According to some scientists agriculture was widespread in the Indian peninsula, 10000–3000 years ago, well beyond the fertile plains of the north. For example, one study reports 12 sites in the southern Indian states of [Tamil Nadu], [Andhra Pradesh]and [Karnataka] providing clear evidence of agriculture of pulses [Vigna radiata] and [Macrotyloma uniflorum], millet-grasses (Brachiaria ramosa and Setaria verticillata), wheats (Triticum dicoccum, Triticum durum/aestivum), barley (Hordeum vulgare), hyacinth bean (Lablab purpureus), pearl millet (Pennisetum glaucum), finger millet (Eleusine coracana), cotton (Gossypium sp.), linseed (Linum sp.), as well as gathered fruits of Ziziphus and two Cucurbitaceae.

Some claim Indian agriculture began by 9000 BC as a result of early cultivation of plants, and domestication of crops and animals. Settled life soon followed with implements and techniques being developed for agriculture. Double monsoons led to two harvests being reaped in one year. Indian products soon reached trading networks and foreign crops were introduced. Plants and animals—considered essential to survival by the Indians—came to be worshiped and venerated.

The middle ages saw irrigation channels reach a new level of sophistication, and Indian crops affected the economies of other regions of the world under Islamic patronage. Land and water management systems were developed with an aim of providing uniform growth.

Despite some stagnation during the later modern era the independent Republic of India was able to develop a comprehensive agricultural programme.