3.3 Dealing with integer and floating point numbers

• Doing math with integer is always safe

.Machine$integer.max 

## [1] 2147483647

# but you can still overflow:
# note: it is a warning not an error 
.Machine$integer.max + 1L 

## Warning in .Machine$integer.max + 1L: NAs produced by integer overflow

## [1] NA

• Floating point number are “limited”: 8 bytes

print(0.1234567891123456789012, digits = 22)

## [1] 0.1234567891123456856439

Hence:

0.1 + 0.2 == 0.3

## [1] FALSE

What can we do!

• round is an usual trick

• the author likes abs(x - y) < very_small_number

abs(sin(pi) - 0) < 2^-26

## [1] TRUE