logo
Speech Recordings  |  Speech Characteristics  |  Normative Studies  |  Depression  |  Coping Skills  |  Program Package  |  Home
unisig
logo

CSELECT — Selects cases (groups), generate learn/test samples

This program is used to define a subsample of cases for subsequent analyses. Cases are selected through the specification of study-numbers, group-numbers, proband-numbers, recording days, and experimental conditions. A sequence of CSELECT-calls accumulates cases, so that any problem-specific subsample of cases can be assembled by a sequence of single steps. An existing subsample of cases is deleted by means of the parameter RSET.

 
            Specificationlist:     CSELECT
            ------------------------------
            I4 STUD                      0
            I4 GRUP                      0
            I4 PROB                      0<
            I4 ATAG                      1  Default-value
            I4 STAT                      0  Default-value
            I4 VERS                      0  Default-value
            I4 SEX                       0  Default-value
            I4 AMIN                      0  Default-value
            I4 AMAX                      0  Default-value
            I4 PROT                      0  Default-value
            I4 RSET                      1  Default-value
            A8 POPU              Undefined
 
            01 STUD Selects studies
            02 GRUP Selects groups
            03 PROB Selects probands
            04 ATAG Selects recording day
            05 STAT Selects experimental condition
            06 VERS Selects version of recording
            07 SEX  Selects gender of probands
            08 AMIN Specifies age interval: minimum age
            09 AMAX Specifies age interval: maximum age
            10 PROT Controls output to display/printer
            11 RSET Controls concatenation of cases
            12 POPU Label that distinguishes between populations
            13 DEMO Examples that illustrate program function
 
            - STUD = s: Study number
                   = 0: all studies
 
            - GRUP = g: Group number
                   = 0: all groups
 
            - PROB = p1(/p2/p3/..  ) probands p1,p2,..
                   = -/a/-b: proband number range a-b
                   = 0: all probands
 
            - ATAG = a: Recording day
                   = 0: all recording days
 
            - STAT = f: Experimental condition
                   = 0: all experimental conditions
 
            - VERS = v: Version of recording
 
            - SEX  = 1: Males only
                   = 2: Females only
                   = 0: males and females
 
            - AMIN = a: Age interval: minimum age
 
            - AMAX = e: Age interval: maximum age
                   = 0: all ages
 
            - PROT = 0: No print output
                     1: Report of all selected probands
 
            - RSET = 0: Add cases to list
                   = 1: Begin new list
 
            - POPU:     Label to be given to this group of probands (8 char)
 
 
            - DEMO: Select 20-30 years old test persons for analysis
 
           

Example

 
 
            &&START CSELECT=Normative speech study zurich: males (study 600)
             STUD=600,ATAG=1,STAT=2,SEX=1,AMIN=20,AMAX=30,PROT=1,RSET=1
             POPU=MALES
              ?
            &&START CSELECT=Normative speech study zurich: females (study 600)
             SEX=2,RSET=0
             POPU=FEMALES
            &&START PATTERNS=Normative speech study zurich (study 600)
             PROT=1,PLOT=2,LPRT=82,SAVE=0,FRST=3,NSPK=15
 
 
           
voxFig14
Fig. 14: Voice sound characteristics possess a distinct "individuality" that allows one, for example, to identify persons on the phone very quickly without speaking explicitly about the identity of the speaker. In fact, voice sound characteristics have a strong biological component in the range of 80% or higher. The inter-individual differences of voice sound patterns between unrelated subjects can be studied in detail by means of a similarity function that quantifies between-subject similarities and dissimilarities as a function of frequency (here: comparison of two repeated assessments of the same subject at 14-day intervals). See Fig. 13 for a comparison of unrelated subjects.

Partners:
Everis, Spain
ETH, Switzerland
UZH, Switzerland
Freiburg, Germany
MA Systems, UK
Bristol, UK
Xiwrite, Italy
Ultrasis, UK
Jaume, Spain
Valencia, Spain
Lanzhou, China

 

EU-Grant (FP7):
248544

[ Comments and suggestions to Webmaster ] k454910@bli.uzh.ch
[ Home ]
Impressum  |  Acknowledgements