4.35 sddsfdfilter

• description:
• synopsis:

  sddsfdfilter [-pipe[=input][,output]] [inputfile] [outputfile] [-columns=indep-variable[,depen-quantity][,depen-quantity...]] [-exclude=depen-quantity[,depen-quantity]] [-clipFrequencies=[high=number][,low=number]] [-threshold=level=value[,fractional][,start=freq][,end=freq]] [-highpass=start=freq,end=freq] [-lowpass=start=freq,end=freq] [-notch=center=center,flatWidth=width1,fullWidth=width2] [-bandpass=center=center,flatWidth=width1,fullWidth=width2] [-filterFile=filename=filename,frequency=columnName, {real=columnName,imaginary=columnName | magnitude=columnName}]
  [-cascade] [-newColumns] [-differenceColumns]

• switches:
  • -pipe[=input][,output] — The standard SDDS Toolkit pipe option.
  • -columns=indepVariable[,depenQuantity][,depenQuantity...] — Gives the name of the independent variable (typically the time variable) with respect to which filtering is done. If no depenQuantity qualifiers are given, then all numerical columns are filtered; otherwise, only the named columns are filtered.
  • -exclude=depenQuantity[,depenQuantity] — Specifies the names of columns to exclude from time filtering, altering whatever selections are made by the -columns option.
  • -clipFrequencies[=low=frequency][,high=frequency] — Specifies clipping frequencies below a given frequency (low qualifier) and/or above a given frequency (high qualifier). Any frequencies in the signals that are clipped are set to zero.
  • -threshold=level=value[,fractional][,start=freq][,end=freq]] — Specifies a threshold level below which Fourier components are set to zero. If the fractional qualifier is given, the level is interpreted as a fraction of the largest component. The start and end qualifiers allow restricting the frequency range over which the threshold is applied.
  • -highpass=start=freq,end=freq — Specifies a highpass filter. Frequency components below the start value are set to zero, while those above the end value are unaffected. Those in between are multiplied by a value that varies linearly from 0 to 1.
  • -lowpass=start=freq,end=freq — Specifies a lowpass filter. Frequency components above the end value are set to zero, while those below the start value are unaffected. Those in between are multiplied by a value that varies linearly from 0 to 1.
  • -notch=center=center,flatWidth=width1,fullWidth=width2 — Specifies a notch filter centered on a given frequency, attenuating completely within a band width1 wide. Frequencies outside a band width2 are unattenuated. Frequencies between the two widths are attenuated by values varying linearing from 0 to 1 as the frequency becomes more distant from the center frequency.
  • -bandpass=center=center,flatWidth=width1,fullWidth=width2 — Specifies a bandpass filter centered on a given frequency, passing a band width1 wide without attenuation. Frequencies outside a band width2 are completely attenuated. Frequencies between the two widths are attenuated by values varying linearing from 1 to 0 as the frequency becomes more distant from the center frequency.
  • -filterFile=filename=filename,frequency=columnName, {real=columnName,imaginary=columnName | magnitude=columnName} — Specifies a filter explicitly as a function of frequency, using either a single value (simple attenuation) or a real and imaginary value. The name of the column containing frequency values must be given with the frequency qualifier. In addition, one must either give the name of the column containing the magnitude (attenuation factor) or else both the names of the columns containing the real and imaginary components. The function thus specified is interpolated linearly to obtain values at any required frequencies. Fourier components at frequency beyond the range in the file are unaffected.
  • -cascade — By default, if several filters are specified using the above options, their output is added. When the cascade option is given, a new sequence is started, with the original signal as input. The output from all cascades is summed to obtain the final result. For example, if one wants to have the nested effect F4(F3(F2(F1(I0))))) (i.e. multiplier effect), one needs to give the command as following:

    sddsfdfilter ¡input¿ ¡output¿ -column=Index,... F1 -cascade F2 -cascade F3 -cascade F4

    If one wants F1(I0) + F2(I0) + F3(I0) + F4(I0) then one needs to write following command:(The illustration of output is shown in figure)

    sddsfdfilter ¡input¿ ¡output¿ -column=Index,... F1 F2 F3 F4

  • -newColumns — Specifies that new columns be created in the output file to hold the result of the filtering. By default, the filtered data is placed in columns with the same names as those in the input file. Using this option, both the filtered and unfiltered data will appear in the output file. The filtered data will have names of the form columnNameFiltered, where columnName is the name of the source column in the input file.
  • -differenceColumns — Specifies that new columns be created in the output file that contain the difference between each original column and the corresponding filtered column. The new columns have names of the form columnNameDifference, where columnName is the name of the source column in the input file.
• see also:
  • sddsdigfilter
  • sddssmooth
• author: M. Borland, ANL/APS.