i randomly generated some data for n=200 and n=2000 from an integer distribution ranging from 20 to 50 ... in the older graphics mode, here is what i got


Histogram of C1   N = 200

Midpoint        Count
      20            6  ******
      24           24  ************************
      28           26  **************************
      32           27  ***************************
      36           28  ****************************
      40           26  **************************
      44           31  *******************************
      48           24  ************************
      52            8  ********


Histogram of C2   N = 2000
Each * represents 5 observation(s)

Midpoint        Count
      20           71  ***************
      22          113  ***********************
      24          131  ***************************
      26          142  *****************************
      28          131  ***************************
      30          125  *************************
      32          151  *******************************
      34          129  **************************
      36          135  ***************************
      38          124  *************************
      40          145  *****************************
      42          110  **********************
      44          130  **************************
      46          121  *************************
      48          120  ************************
      50          122  *************************

now, i also looked at the pro graphics under each case and found for n=200, there were about 16 intervals ... and when n=2000, there were about 31 intervals ... or about twice ... similar to the difference (in ratio) in the older graphics ...

help in minitab does not provide any clue about the algorithm on this ... but clearly, it treats them differently ... but of course, one can change this to whatever YOU want ... with subcommands

At 11:01 PM 1/4/01 -0500, Jineshwar Singh wrote:
To determine the number of classes for a histogram, Excel uses square root
of the number of observations. Is it also true for the number of
observations greater than 200, say, for 2000?. Does the MINITAB use the same
for determining the number of classes for a histogram?
Any help would be appreciated.





=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
                  http://jse.stat.ncsu.edu/
=================================================================

================================================================= Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =================================================================

Reply via email to