2 4 I. i.7
0 0 0 1 1 2 2
Interval adds an extra count for bin above the last range.
still, just the count as:
2 4 #/.~@:I. i.7
3 2 2
and then with printed intervals,
2 4 ([ (] ,.~ [ {.!._~ #@]) #/.~@:I.) i.7
2 3
4 2
_ 2
2 4 9 ([ (] ,.~ [ {.!._~ #@]) #/.~@:I.) i.7
2 3
4 2
9 2
________________________________
From: Rudolf Sykora <rudolf.syk...@gmail.com>
To: programm...@jsoftware.com
Sent: Thursday, June 15, 2017 7:50 AM
Subject: [Jprogramming] histogram from Essays
Hello,
I found a code to produce a histogram at
http://code.jsoftware.com/wiki/Essays/Histogram.
However, both the description and the function seem odd.
The code reads
histogram=: <: @ (#/.~) @ (i.@#@[ , I.)
1)
They say: ---The left argument is a list of interval *start* points.---,
which I think is not quite right (or is at least deceptive):
2 3 histogram 1
1 0
2 3 histogram 2
1 0
(thus x are actually inclusive endpoints).
2)
Furthemore and more importantly:
2 histogram 1
1
2 histogram 10
0 0
which (the 2nd case) is not nice, since the shape of the result
changes depending on data *value*.
I'd rather have a verb that would output
2 hist 1
1 0
2 hist 10
0 1
Does anyone have a verb like that? (I will try to design one myself, but
am asking anyway.)
Thanks
Ruda
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
----------------------------------------------------------------------
For information about J forums see http://www.jsoftware.com/forums.htm
