B. Wilson,
I like your last approach:
, 10#.,"1/^:3~ ,. 0 1 2 6
0 1 2 6 10 11 12 16 20 21 22 26 60 61 62 66 100 101 102 106 110 111 112 116
120 121 122 126 160 161 162 166 200 201 202 206 210 211 212 216 220 221 222
226 260 261 262 266 600 601 602 606 610 611 612 616 620 621 622 626 660 661
662
Over lunch, I just realized that my second implementation using integer digits
could be expressed a lot more directly:
, 10#. ,"1/^:3~ ,. 0 1 2 6
0 1 2 6 10 11 12 16 20 21 22 26 60 61 62 66 100 101 102 106 110 111 ...
Which nicely mirrors the implementation using strings.
On Mon, O
I was a bit eager on the send button with that last message.
Here is the previous approach, removing some superfluous stuff:
, ". ,"1/^:3~ ,. '0126'
0 1 2 6 10 11 12 16 20 21 22 26 60 61 62 66 100 101 102 106 110 111 ...
And here is essentially the same approach, but treating digit
What about something like this?
, ". (,/@:,"1/)^:3~ ,.'0126'
0 1 2 6 10 11 12 16 20 21 22 26 60 61 62 66 100 101 102 106 110 111 ...
If S is the set of our digits, then the above essentially is just the Cartesian
product S^4.
On Sun, Oct 20, 2019 at 06:08:48PM -0500, Skip Cave wrote
Is there a less-verbose way to list all the integers (one-digit, two-digit,
three-digit, four-digit) can be formed by using the characters 0, 1, 2, and
6 once? I used multiple takes:
*#b=.~.>10#.ea({1{."1 a),({2{."1 a),({3{."1 a),{a=.(perm 4){0 1 2 6*
*49*
*b*
*0 1 2 6 10 12 16 20 21 26 60 61 6
A new version of jmf addon is available. It provides new features that will
eventually be used in Jd. It should be upward compatible but it is
complicated so use care and be on the look out for problems.
See doc_jmf_ for a usage summary.
---
