Ntuple creation

Creation of an Row-Wise Ntuple (RWN) and first look at its contents
1     NTUPLE/CREATE 10 'CERN Population' 11 ' ' 3500  _                                        
      Category Division Flag Age Service Children Grade _
      Step Nation Hrweek Cost
      *
2     NTUPLE/READ 10 APTUPLE.DAT                                                               
3     HISTO/FILE 1 RWN_APTUPLE.HBOOK 1024 N                                                    
3     HROUT 10                                                                                 
4     NTUPLE/PRINT 10                                                                          
      zone 1 2
6     OPT STAT                                                                                 
6     SET STAT 110                                                                             
4     NTUPLE/PLOT 10.Age                                                                       
      ntuple/plot 10.Division

/NTUPLE/CREATE IDN TITLE NVAR CHRZPA NPRIME VARLIST : Allows to create an Ntuple. An Ntuple is a matrix of n columns. Each line of the matrix is often called an ``event''. Internally there is two different way to access the data: by rows (Row-Wise Ntuple) or by columns (Column-Wise Ntuple). The Ntuple may be created either in memory or, if necessary, using an automatic overflow to an histogram file.
NT/READ allows to fill an RW/Ntuple with numeric values read from an existing ASCII file.
Like histograms, Ntuples are HBOOK objects and can be stored into histogram files opened via the command HIST/FILE .
The command NT/PRINT gives the description of the Ntuple (see next page).
NT/PLOT allows to plot an Ntuple. The syntax is:
```
                      NT/PLOT nid.n ......
```
where `` nid '' is the Ntuple identifier (a number) and `` n '' is the number or the name of one of the variable in the Ntuple. By default, if `` n '' is not specified, the first variable of the Ntuple is ploted.
Note also:
OPT STAT and SET STAT are used to plot some statistical informations.

Creation of Column-Wise Ntuple (CWN)
1     HISTO/FILE 1 CWN_APTUPLE.HBOOK 1024 N                                                    
2     CALL CERNPOP.F                                                                           
      hrout 11
      ntuple/print 11
      zone 1 2
      opt stat
      set stat 110
      ntuple/plot 11.Age
3     NTUPLE/PLOT 11.Division

A new HBOOK file is open. If the Ntuple created after doesn't fit in memory, it will be automatically write on this file.
This command create and read a CW/Ntuple. It is the equivalent of the /NTUPLE/CREATE and /NTUPLE/READ commands in the previous example (for the time being these commands work only with the RWN format). For more details on the CW/Nutples management see the HBOOK manual. The main advantages of the CW/Nutples compare to the RW/Nutples are:
1. The speed: only the used data are read.
2. The possibility to have typed variable (not only real).
3. The compactness.
4. The ``Array variables'' with a fixed or variable length.
The axis are directly drawn with character labels.

COMIS routine used to create a CW/Ntuple
      Subroutine cernpop
*
      integer category, flag, age, service, children, grade, step,
     +        hrweek, cost
      common /cern/ category, flag, age, service, children, grade,
     +              step, hrweek, cost
      character*4   division, nation
      common /cernc/ division, nation
*
      character*132 chform
      dimension    rdata(11)
      character*4 divs(13), nats(15)
      data divs /'AG', 'DD', 'DG', 'EF', 'EP', 'FI', 'LEP', 'PE',
     +           'PS', 'SPS', 'ST', 'TH', 'TIS'/
      data nats /'AT', 'BE', 'CH', 'DE', 'DK', 'ES', 'FR', 'GB',
     +           'GR', 'IT', 'NL', 'NO', 'PT', 'SE', 'ZZ'/
*
      open(unit=41,file='aptuple.dat',status='old')
*
      call hbnt(11,'CERN Population (CWN)',' ')
      chform = ' CATEGORY[100,600]:I, FLAG:U:6, AGE[1,100]:I,'//
     +         ' SERVICE[0,60]:I, CHILDREN[0,10]:I, GRADE[3,14]:I,'//
     +         ' STEP[0,15]:I, HRWEEK:I, COST:I'
      call hbname(11, 'CERN', category, chform)
      chform = 'DIVISION:C,           NATION:C'
      call hbnamc(11, 'CERN', division, chform)
*
10    read(41, '(10F4.0, F7.0)', end=20) rdata
      category = rdata(1)
      division = divs(int(rdata(2)))
      flag     = rdata(3)
      age      = rdata(4)
      service  = rdata(5)
      children = rdata(6)
      grade    = rdata(7)
      step     = rdata(8)
      nation   = nats(int(rdata(9)))
      hrweek   = rdata(10)
      cost     = rdata(11)
      call hfnt(11)
      goto 10
*
  20  close (41)
      end

RWN NT/PRINT output
             ********************************************************
             * NTUPLE ID=   10  ENTRIES=   3354   CERN Population   *
             ********************************************************
             *  Var numb  *   Name    *    Lower     *    Upper     *
             ********************************************************
             *      1     * CATEGORY  * 0.102000E+03 * 0.567000E+03 *
             *      2     * DIVISION  * 0.100000E+01 * 0.130000E+02 *
             *      3     * FLAG      * 0.000000E+00 * 0.310000E+02 *
             *      4     * AGE       * 0.210000E+02 * 0.640000E+02 *
             *      5     * SERVICE   * 0.000000E+00 * 0.350000E+02 *
             *      6     * CHILDREN  * 0.000000E+00 * 0.600000E+01 *
             *      7     * GRADE     * 0.300000E+01 * 0.140000E+02 *
             *      8     * STEP      * 0.000000E+00 * 0.150000E+02 *
             *      9     * NATION    * 0.100000E+01 * 0.150000E+02 *
             *     10     * HRWEEK    * 0.200000E+01 * 0.440000E+02 *
             *     11     * COST      * 0.686000E+03 * 0.188530E+05 *
             ********************************************************

CWN NT/PRINT output
        ******************************************************************
        * Ntuple ID = 11     Entries = 3354      CERN Population (CWN)
        ******************************************************************
        * Var numb * Type * Packing *    Range     *  Block   *  Name    *
        ******************************************************************
        *      1   * I*4  *    11   * [100,600]    * CERN     * CATEGORY
        *      2   * U*4  *    6    *              * CERN     * FLAG
        *      3   * I*4  *    8    * [1,100]      * CERN     * AGE
        *      4   * I*4  *    7    * [0,60]       * CERN     * SERVICE
        *      5   * I*4  *    5    * [0,10]       * CERN     * CHILDREN
        *      6   * I*4  *    5    * [3,14]       * CERN     * GRADE
        *      7   * I*4  *    5    * [0,15]       * CERN     * STEP
        *      8   * I*4  *         *              * CERN     * HRWEEK
        *      9   * I*4  *         *              * CERN     * COST
        *     10   * C*4  *         *              * CERN     * DIVISION
        *     11   * C*4  *         *              * CERN     * NATION
        ******************************************************************
        *  Block   * Unpacked Bytes * Packed Bytes *   Packing Factor    *
        ******************************************************************
        * CERN     *    44          *    22        *    2.000            *
        * Total    *    44          *    22        *    2.000            *
        ******************************************************************
        * Number of blocks = 1     Number of columns = 11                *
        ******************************************************************

			PAW Tutorial	Last update: 1999/04/01 14.58	Olivier.Couet@Cern.Ch