For most usual applications, these parameters are automatically generated from the button “Create a command file" of the main menu. However, the command file could also be edited in a subsequent phase (change of parameters values, replacement of some default values, addition of a new step, etc.)
Parameters naxe: number of requested axes. nvedi: number of eigenvalues (histogram). ledin: printing the coordinates of the individuals (0=no, 1=act, 2=tot). naxed: number of printed coordinates. niter: number of iterations (case of iterative algorithm) (0=no). nictl: number of "controlled iterations" (0=no). nbase: dimension of the approximation subspace. nsimu: number of bootstrap replication (0 = no bootstrap, max = 30) nboot: type of bootstrap replication ( 0 = "total bootstrap", 1= partial bootstrap [default]) ndata: printing the replicated data tables (1 = yes, 0 = no [default])
Parameters naxe: number of principal coordinates. ledit: parameter for printing the rows (0=no, 1=yes). ngraf: number of requested graphics npage: number of pages of the graphics nrows: number of lines of the graphics nsimu: number of bootstrap replication (0 = no bootstrap, max = 30) nboot: type of bootstrap replication 0 = "total bootstrap", [bootstrapping the words] 1 = partial bootstrap [default] [bootstrapping the words] 2 = does not apply here 3 = total specific bootstrap [bootstrapping the respondents] 4 = partial specific bootstrap [bootstrapping the respondents] ndata: printing the replicated data tables (1 = yes, 0 = no [default])
Parameters nqexa: number of variables in the dictionary. nxmod: maximal number of categories for a variable. ledit: printing mode of the dictionary (0=no, 1=short, 2=long) nidi: number of groups of 4 characters (identifier of individuals) (0=no). niexa: number of individuals in file ndonz. test: value given to missing numerical data (default=.999). nedit: number of individuals in a check printout (0=no).
Parameters ityp: type of textual data file ntexz (tex=1, enq =2) nbqt: number of questions per individual nlig: number of responses to be printed for each text (0 = only the titles) lirep: indicates implicit end of a response (1= 1 line (row)/1 response)
Parameters naxe: number of requested principal coordinates ledit: printing the responses (0 = no, 1 = coor, 2 = tot) ngraf: number of requested graphics npage: number of pages of the graphics nrows: number of lines of the graphics nbase: dimension of the approximation space (0= main core diagonalization) niter: number of iterations (if nbase > 0)
Parameters naxe: number of described axes lign: simultaneous listing of the rows (0=no, 1=yes). nmax: number of printed "extreme elements" for each principal axis
Parameters ledit: printing identifiers of individuals (0 = no printing, 1 = identifiers of respondents are printed, default = 0) lspb: use of a file nspb (0=no, 1=yes) --- key-word of headings : forme end
Parameters ledit: printing of words (0=no, 1=nspc, 2=tot). lclas: printing sorted words (0=no, 1=yes). --- key-word of headings : delet equiv end
Parameters ednom: printing the tables (classes * questions) (0=no). lpnom: describing partition with questions (0=no, 1=yes). cnomi: describing classes with questions (0=no). pmoda: describing partition with categories (0=no). cmoda: describing classes with categories (0=no). pcmin: minimum relative weight (%) for a category. lsupr: characteristic category if %(cat./class) > %(cat./total) (0=no, 1=yes). pfreq: describing partition with frequencies (0=no). lpfre: printing thetable (classes * frequencies) (0=no, 1=act, 2=tot). cfreq: describing classes with frequencies (0=no). edcon: describing partition with numerical variables (0=no, 1=yes). ccont: describing classes with numerical variables (0=no).
Parameters ednom: printing the tables (classes * questions) (0=no). lpnom: describing partition with questions (0=no, 1=yes). cnomi: describing classes with questions (0=no). pmoda: describing partition with categories (0=no). cmoda: describing classes with categories (0=no). pcmin: minimum relative ( % ) weight for a category. lsupr: characteristic category if %(cat./class) > %(cat./total)(0=no,1=yes). pfreq: describing partition with frequencies (0=no). lpfre: printing the table (classes * frequencies) (0=no, 1=act, 2=tot). cfreq: describing classes with frequencies (0=no). edcon: describing partition with numerical variables (0=no, 1=yes). ccont: describing classes with numerical variables (0=no). paxe: describing partition with axes (0=no). caxe: describing classes with axes(0=no). ledin: printing correspondence classes*individuals (0=no,1=compo, 2=affec, 3=tot).
Parameters seuil: threshold for selecting most significant items. lcrim: criterion of selection for the categories ( coordinate or test-value) vtmin: test-value threshold for selecting categories.
Parameters niter: Maximum number of iterations acdif: Maximum difference between required and obtained percentages liste: Printing the individual weights (0=no, 1=yes). ledit: Printing the future provisional dictionnary (0=no , 1=short, 2=long). nxvrd: Maximal number of categorical variables used for the iterative fitting key word: end (after lists of target [or: required] percentages)
Parameters nomot: number of charac. words for each text. norep: number of charac. responses for each text.
Parameters nvsel: index of the categorical variable defining the groupings of texts ledit: parameter for printing the table words*texts (0=no, 1=yes).
Parameters naxe: number of requested axes. pcmin: threshold for "cleaning" of the active categories (in percent). lburt: printing the Burt table (0=no, 1=mass, 2=tot, 3=prof). nvedi: number of eigenvalues (histogram). naxed: number of printed coordinates. ledin: printing the coordinates of individuals (0=no, 1=act, 2=tot). ledco: printing the correlations variable-axers (0=no, 1=yes). nsimu: number of bootstrap replication (1 = no bootstrap) nboot: (1 = partial bootstrap [default]) (0 = total bootstrap. In such a case, the "cleaning of rare categories" [parameter pcmin above] is not allowed)
Parameters coef: technical coefficient (frequencies of word-lengths) (default value = 1; put coef=10 for large corpora). nseu: frequency threshold of the kept words nxlet: maximum number of letters per word. nxmax: maximum number of distinct words. nxsig: maximum number of first letters of word. ledit: printing the words (0=no, 1=alpha, 2=tri, 3=tot). nmomi: threshold for the length of words --- key-word of headings weak strong accent histo end
Parameters nxpar: maximum number of requested dendrogram cut. niter: number of "consolidation" iterations (0=no). ledin: printing the correspondence classes-individuals (0=no; 1=class-->individuals; 2=individuals-->classes; 3= both 1 and 2). naxed: number of axes for printed test-values (0=no).
Parameters and individuals (rows) ) ledit: printing the number of permuted elements (0=no, 1=yes).
Parameters ngraf: number of requested graphics npage: number of pages of the graphics nrows: number of lines of the graphics
Parameters naxe: number of printed axes ngraf: number of requested graphics npage: number of pages for the graphics nrows: number of lines for the graphics zoom: maximum distance to the origin (as numbers of standard deviations)
Parameters lcorr: analysis of the correlation matrix (0=no, 1=yes). naxe: number of requested axes nvedi: number of eigenvalues (histogram). ledin: printing the coordinates of the individuals (0=no, 1=act, 2=tot). naxed: number of printed coordinates. navir: number of digits (right to the dot) (printing of the coordinates). nsimu: number of bootstrap replication (0 = no bootstrap) nboot: type of bootstrap replication (0 = "total bootstrap", 1= partial bootstrap, 2 = bootstrap on variables)
Parameter norep: number of characteristic responses for each text
Parameters naxu: number of kept axes. nterm: number of kept terminal elements of tree. nived: number of levels in histogram of indices (0=no). lterm: printing the composition of terminal elements (0=no, 1=yes). naxed: number of coordinates for printing terminal elements (0=no). ldesc: describing the nodes of the tree (0=no, 1=yes). ldend: printing the dendrogram (0=no, 1=dense, 2=large).
Parameters lsupr: missing value ( =0 : individuals with m.v. are kept) ( =1 : individuals with m.v. discarded) ( default = 1) lstat: printing statistics (=1) or not (default=0) lmat: printing covariance matrix (default = 0 = no) (1 = covariance, 2 = correlation) Then: regression formulae: e.g.: V8 = V9 + V57 (one per line) (V8, V9, V57 = continuous variables) V8 = V67 + V12*V15 (V8 = continuous, V67, V12, V15 = categorical) Ending the commands with the key word "end"
Parameters nxlon: maximum length of the segment nxseg: maximum number of requested segments nseug: frequency threshold of segments( general) nseu2: frequency threshold of segments of length 2 nseu3: frequency threshold of segments of length 3 nxle2: maximum number of segments of length 2
Parameters lseli: mode of selection of individuals (0=all, 1=list, 2=filt). imass: index of variable 'weight of individuals' (0=uniform weights). lzero: coding missing responses (0=norec, 1=rec). ledit: printing dictionary of selected variables (0=no, 1=short,2=long). nxfil: maximum number of constants used in filter (lseli = filt ou 2). The selection commands use following key words : act ill end nomi cont freq
Parameters numq: index of the selected question if numq = -1 or numq = list : several questions will be merged (list of questions numbers must follow next line) lseli: mode of selection of the individuals (0=tot, 1=list, 2=filt). nxfil: maximum number of constantes used in filter (lseli = filt or 2). ldona: presence of files ndona and ndica (files produced by step ardat)
Parameters naxu: number of kept axes. niter: number of iterations for each basic partition. lmeth: type of clustering (1=mobil,2=stabl). seuil: number of kept terminal elements lresi: building of a resisual cluster (1=yes,0=no). nived: number of levels in the histogram of indices (0=no). ldesc: describing nodes of the tree (0=no, 1=yes). ldend: printing dendrogram (0=no, 1=dense, 2=large).
Parameters nseu: threshold of frequency for selecting words. nmomi: minimum number of letters of a kept word. nremi: minimum number of words of a kept response. nxmsu: maximum number of words explicitly eliminated. ledit: printing the dictionaries (0=no, 1=new, 2=tot).
Parameters ledit: printing identifiers of supplementary texts. --- key-word of headings forme text end
Parameters lcont: printing elementary statistics about numerical variables (0=no, 1=yes). lhist: printing the histograms of numerical variables (0=no, 1=yes). unit: default unit for histograms. zmin: minimal default lower bound for histograms. zmax: maximal default lower bound for histograms. nclas: default number of classes for histograms. lcom: commands for numerical variables (0=no, 1=yes). nxvdi: number maximum of values for a discretized numerical variable. --- key-words of commands zmin zmax unit nclas hist libre disc end
Parameters imass: index of the variable 'weight of individuals' (0=uniform weight generated = default). lzero: coding missing responses (0= norec = missing values ignored - default). (1 = rec = m.v. become a new category) lfilt: use of a filter (0 = default = no filter) (1 = a filter will follow) ledic: printing variables dictionary (0 = default = no, 1 = short, 2 = large) letyp: printing standard deviation (case of a table of mean values) (0 = no = default) (1 = yes: means and std dev. are printed) lnull: printing columns whose weight is zero (0 = no, 1 = yes = default)
Parameter ledit: printing the table words*texts (0=no, 1=yes).