Errors Accuracy and Roundoff

  1. Binary
    1. Convert binary to deciman (>1 and <1)
  2. Floating point representation
    1. why is it call that (float point representation)?
    2. how are 1.001E3 and 2.2E-1 added?
    3. What are the parts of a floating point number including their names?
  3. Precision
    1. What is the relative error (machine precision) for doubles on a computer?
    2. How many bits are in a byte?
    3. How many bytes and bits are in a double?
    4. In math, we have $\mathbb{N,,Z,,Q,,R,,C}$. Which do we have naturally on computers?
    5. Is $a+b+c$ the same as $a+c+b$? Why?
  4. Roundoff error analysis
    1. What is 1/10 in binary? (Why is that interesting?)
    2. Whis is best: $a^2-b^2$ or $(a-b)(a+b)$?
    3. Which calculation should make you more nervous and why: $\prod_ix_i$, $\sum_ix_i$?
    4. Why is subtracting two nearly equal numbers bad?