Problem 35
Question
. Did you ever wonder how people find the decimal expansion of \(\pi\) to a large number of places? One method depends on the following identity (see Problem 34 of Section 1.9). $$ \pi=16 \tan ^{-1}\left(\frac{1}{5}\right)-4 \tan ^{-1}\left(\frac{1}{239}\right) $$ Find the first 6 digits of \(\pi\) using this identity and the series for \(\tan ^{-1} x\). (You will need terms through \(x^{9} / 9\) for \(\tan ^{-1}\left(\frac{1}{5}\right)\), but only the first term for \(\left.\tan ^{-1}(1 / 239) .\right)\) In 1706, John Machin used this method to calculate the first 100 digits of \(\pi\), while in 1973 , Jean Guilloud and Martine Bouyer found the first 1 million digits using the related identity $$ \pi=48 \tan ^{-1}\left(\frac{1}{18}\right)+32 \tan ^{-1}\left(\frac{1}{57}\right)-20 \tan ^{-1}\left(\frac{1}{239}\right) $$ In \(1983, \pi\) was calculated to over 16 million digits by a somewha different method. Of course, computers were used in these recen calculations.