-
Notifications
You must be signed in to change notification settings - Fork 38
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
0x10000000000000000.pow(3) is computed incorrectly #30
Comments
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 22, 2015
Merged
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 23, 2015
Something like `(odd * 2**64)**3` should be computed as `odd**3 * 2**(64 * 3)`, but it was being computed as just `odd**exp * 2**64`. Fixes Aatch#30.
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 23, 2015
Something like `(odd * 2**64)**3` should be computed as `odd**3 * 2**(64 * 3)`, but it was being computed as just `odd**3 * 2**64`. Fixes Aatch#30.
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 23, 2015
Bias toward long sequences of just 0's or just 1's. GMP's mpz_rrandomb's docs say that these seem to be more likely to pick up edge cases. This assertion make sense: for most algorithms, only zero or all-bits-set (or 1) have useful special cases. Also, this seems to be bourne out in practice with Ramp e.g. Aatch#30 was only picked up by a earlier version of biasing like this, with uniform bits, the chance of getting a trailing zero word to trigger that is 2**-64. Examples of numbers generated by this distribution (in hex): bdfffffffffffffffffffffffffffffffffffffffffffffffffae00000000000000000000000000000000000000000a3a188 bfffffffffffffc2222000000000000000000000000000000000000000000000000000000000000f9a00000000bd4f845457 b9fffffffffff25000097e8126ffffffffffefffffffffffffffffffffffffffffffffffffffffd5c7aeffffffffffffffff f00000000000003ffffffffff1f14fffffffffffff3fffffff76dd6fffffffffffffffffffffffffff1288b9000000000000 c6cff00000000007ff8ffffffffffffff63ad6dce80000000000000000000000000000020000000000000000000000f3ea39 4fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffcd996c00000000000000000000000000 c00000000000000000003fffffff100000000000000000000000000000000020000000000000000000000000033ec4958000 bae4a67fe0cffffffffff70bcd7c80220000000000000000000000000000000000000000840008d9a9aa009f400000000000 beacc705c1fffffff3a8c2d9788d3ca516322fffffffffffffffffffffffe0805c5da9b071fffffffffffffddfffffffffff ef23600000000000000000b54d000000000000000000000000000000000000a5fffffffffd83200000000000000000000000 7ffffffffffffffffffffffffffffffffffffffffffffffffffffffff7cffffffffffcfffffffffffff9aefffffffffffe7f
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 24, 2015
Bias toward long sequences of just 0's or just 1's. GMP's mpz_rrandomb's docs say that these seem to be more likely to pick up edge cases. This assertion make sense: for most algorithms, only zero or all-bits-set (or 1) have useful special cases. Also, this seems to be bourne out in practice with Ramp e.g. Aatch#30 was only picked up by a earlier version of biasing like this, with uniform bits, the chance of getting a trailing zero word to trigger that is 2**-64. Examples of numbers generated by this distribution (in hex): bdfffffffffffffffffffffffffffffffffffffffffffffffffae00000000000000000000000000000000000000000a3a188 bfffffffffffffc2222000000000000000000000000000000000000000000000000000000000000f9a00000000bd4f845457 b9fffffffffff25000097e8126ffffffffffefffffffffffffffffffffffffffffffffffffffffd5c7aeffffffffffffffff f00000000000003ffffffffff1f14fffffffffffff3fffffff76dd6fffffffffffffffffffffffffff1288b9000000000000 c6cff00000000007ff8ffffffffffffff63ad6dce80000000000000000000000000000020000000000000000000000f3ea39 4fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffcd996c00000000000000000000000000 c00000000000000000003fffffff100000000000000000000000000000000020000000000000000000000000033ec4958000 bae4a67fe0cffffffffff70bcd7c80220000000000000000000000000000000000000000840008d9a9aa009f400000000000 beacc705c1fffffff3a8c2d9788d3ca516322fffffffffffffffffffffffe0805c5da9b071fffffffffffffddfffffffffff ef23600000000000000000b54d000000000000000000000000000000000000a5fffffffffd83200000000000000000000000 7ffffffffffffffffffffffffffffffffffffffffffffffffffffffff7cffffffffffcfffffffffffff9aefffffffffffe7f
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 26, 2015
Bias toward long sequences of just 0's or just 1's. GMP's mpz_rrandomb's docs say that these seem to be more likely to pick up edge cases. This assertion make sense: for most algorithms, only zero or all-bits-set (or 1) have useful special cases. Also, this seems to be bourne out in practice with Ramp e.g. Aatch#30 was only picked up by a earlier version of biasing like this, with uniform bits, the chance of getting a trailing zero word to trigger that is 2**-64. Examples of numbers generated by this distribution (in hex): bdfffffffffffffffffffffffffffffffffffffffffffffffffae00000000000000000000000000000000000000000a3a188 bfffffffffffffc2222000000000000000000000000000000000000000000000000000000000000f9a00000000bd4f845457 b9fffffffffff25000097e8126ffffffffffefffffffffffffffffffffffffffffffffffffffffd5c7aeffffffffffffffff f00000000000003ffffffffff1f14fffffffffffff3fffffff76dd6fffffffffffffffffffffffffff1288b9000000000000 c6cff00000000007ff8ffffffffffffff63ad6dce80000000000000000000000000000020000000000000000000000f3ea39 4fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffcd996c00000000000000000000000000 c00000000000000000003fffffff100000000000000000000000000000000020000000000000000000000000033ec4958000 bae4a67fe0cffffffffff70bcd7c80220000000000000000000000000000000000000000840008d9a9aa009f400000000000 beacc705c1fffffff3a8c2d9788d3ca516322fffffffffffffffffffffffe0805c5da9b071fffffffffffffddfffffffffff ef23600000000000000000b54d000000000000000000000000000000000000a5fffffffffd83200000000000000000000000 7ffffffffffffffffffffffffffffffffffffffffffffffffffffffff7cffffffffffcfffffffffffff9aefffffffffffe7f
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 27, 2015
Bias toward long sequences of just 0's or just 1's. GMP's mpz_rrandomb's docs say that these seem to be more likely to pick up edge cases. This assertion make sense: for most algorithms, only zero or all-bits-set (or 1) have useful special cases. Also, this seems to be bourne out in practice with Ramp e.g. Aatch#30 was only picked up by a earlier version of biasing like this, with uniform bits, the chance of getting a trailing zero word to trigger that is 2**-64. Examples of numbers generated by this distribution (in hex): bdfffffffffffffffffffffffffffffffffffffffffffffffffae00000000000000000000000000000000000000000a3a188 bfffffffffffffc2222000000000000000000000000000000000000000000000000000000000000f9a00000000bd4f845457 b9fffffffffff25000097e8126ffffffffffefffffffffffffffffffffffffffffffffffffffffd5c7aeffffffffffffffff f00000000000003ffffffffff1f14fffffffffffff3fffffff76dd6fffffffffffffffffffffffffff1288b9000000000000 c6cff00000000007ff8ffffffffffffff63ad6dce80000000000000000000000000000020000000000000000000000f3ea39 4fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffcd996c00000000000000000000000000 c00000000000000000003fffffff100000000000000000000000000000000020000000000000000000000000033ec4958000 bae4a67fe0cffffffffff70bcd7c80220000000000000000000000000000000000000000840008d9a9aa009f400000000000 beacc705c1fffffff3a8c2d9788d3ca516322fffffffffffffffffffffffe0805c5da9b071fffffffffffffddfffffffffff ef23600000000000000000b54d000000000000000000000000000000000000a5fffffffffd83200000000000000000000000 7ffffffffffffffffffffffffffffffffffffffffffffffffffffffff7cffffffffffcfffffffffffff9aefffffffffffe7f
huonw
added a commit
to huonw/ramp
that referenced
this issue
Oct 27, 2015
Bias toward long sequences of just 0's or just 1's. GMP's mpz_rrandomb's docs say that these seem to be more likely to pick up edge cases. This assertion make sense: for most algorithms, only zero or all-bits-set (or 1) have useful special cases. Also, this seems to be bourne out in practice with Ramp e.g. Aatch#30 was only picked up by a earlier version of biasing like this, with uniform bits, the chance of getting a trailing zero word to trigger that is 2**-64. Examples of numbers generated by this distribution (in hex): bdfffffffffffffffffffffffffffffffffffffffffffffffffae00000000000000000000000000000000000000000a3a188 bfffffffffffffc2222000000000000000000000000000000000000000000000000000000000000f9a00000000bd4f845457 b9fffffffffff25000097e8126ffffffffffefffffffffffffffffffffffffffffffffffffffffd5c7aeffffffffffffffff f00000000000003ffffffffff1f14fffffffffffff3fffffff76dd6fffffffffffffffffffffffffff1288b9000000000000 c6cff00000000007ff8ffffffffffffff63ad6dce80000000000000000000000000000020000000000000000000000f3ea39 4fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffcd996c00000000000000000000000000 c00000000000000000003fffffff100000000000000000000000000000000020000000000000000000000000033ec4958000 bae4a67fe0cffffffffff70bcd7c80220000000000000000000000000000000000000000840008d9a9aa009f400000000000 beacc705c1fffffff3a8c2d9788d3ca516322fffffffffffffffffffffffe0805c5da9b071fffffffffffffddfffffffffff ef23600000000000000000b54d000000000000000000000000000000000000a5fffffffffd83200000000000000000000000 7ffffffffffffffffffffffffffffffffffffffffffffffffffffffff7cffffffffffcfffffffffffff9aefffffffffffe7f
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
There has to be at least that many trailing zeros, presumably so that the lowest word is empty. And
.pow(2)
is computed correctly.The text was updated successfully, but these errors were encountered: