stata 编程教程(PPT).pdfVIP

14.170: Programming for Economists 1/12/2009-1/16/2009 Melissa Dell Matt Notowidigdo Paul Schrimpf Lecture 5, Large Data Sets in Stata + Numerical Precision Overview • This lecture is part wrap-up lecture, part “tips and tricks” • Focus is on dealing with large data sets and on numerical precision • Numerical precision – Introduction to binary representation – Equilibrating matrices • Large data sets – How Stata represents data in memory – Speeding up code – Tips and tricks for large data sets Numerical precision • What the @*%$!^ is going on here? local a = 0.7 + 0.1 local b = 0.8 display (`a == `b) local a = 0.75 + 0.05 local b = 0.8 display (`a == `b) Binary numbers • Computers store numbers in base 2 (“bits”) 1410 = 11102 (14 = 2 + 4 + 8) 17010 = 101010102 (170 = 2 + 8 + 32 + 128) How are decimals stored? Binary numbers, con’t 0.875 = 0.111 10 2 (0.875 = 0.5 + 0.25 + 0.125) 0.80 = 0.110011001100 10 2 0.70 = 0.101100110011 10 2 0.10 = 0.000110011001 10 2 0.75 = 0.11 10 2 0.05 = 0.000011001100 10 2 QUESTION: Is there a repeating decimal in base 10 that is not repeating in base 2? Precision issues in Mata mata A = (1e10, 2e10 \ 2e-10, 3e-10) A rank (A) luinv (A) A_inv = (-3e-10, 2e10 \ 2e-10, -1e10) I = A * A_inv I end Precision issues in Mata Precision issues in Mata Mata r = c =