What's the fastest way to do this..:?
input: bytestream of "0123456789"
output: integer array
input: integer array
output: bytestream of "0123456789"
Code:
INPUT a
DO
a1$ = LTRIM$(STR$(a MOD 2)) + a1$
a = INT(a / 2)
LOOP UNTIL a <= 0
PRINT a1$
remander method
Code:
INPUT a
a$ = ""
DO
if a and 1 then a$ = a$ + "1" else a$ = a$ + "0"
a = a \ 2
LOOP UNTIL a = 0
PRINT a$
a little bit faster. It's the fastest way I know anyways. I might have mixed up the 4th line though. Untested.
I use the same basic algorithm, the idea to optimize is to use as fewest string operations, (concatenate strings is an expensive job).
Code:
input a%
if a% = 0 then
a$ = "0"
else
' in i% how many digits i need
i%=0
b% = a%
while b% > 0
i% = i% + 1
b% = b% \ 2
wend
' i make an string with "0"s
a$ = string$(i%, "0")
i% = 1
while a% > 0
' just replace the 1
if a% and 1 then
mid$(a$,i%,1) = "1"
endif
a% = a% \ 2
wend
endif
The code surely will not be very fast with small numbers.
I want to use integer arrays...............................................................................................................................................................................................................................
Ok, let me try:
For one way:
Code:
number& = 6445
REDIM Array(LOG(number&) / LOG(2)) AS INTEGER
FOR I = 0 TO UBOUND(Array)
IF (2& ^ I ) AND number& THEN Array(I) = 1 ELSE Array(I) = 0
NEXT I
For the other:
Code:
'the array is called 'array'
FOR I = 0 TO UBOUND(Array)
IF Array(I) THEN nr& = nr& + (2& ^ I)
NEXT I
I hope this suits you
Give another example, so i can understand what are you expecting.
neo has the right idea, but he's using a long instead of a string.
Example:
"89865876578" becomes an integer array that uses as many bits as the minimum bits required to hold this string's number, 16 bits per integer. (although 15 would be OK...)
the integer array becomes "89865876578"