aststatistics - Man Page

calculate statistics of a dataset

Statistics is part of GNU Astronomy Utilities 0.23. Statistics will do statistical analysis on the input dataset (table column or image). All blank pixels or pixels outside of the given range are ignored. You can either directly ask for certain statistics in one line/row as shown below with the same order as requested, or get tables of different statistical measures like the histogram, cumulative frequency style and etc. If no particular statistic is requested, some basic information about the dataset is printed on the command-line.

For more information, please run any of the following commands. In particular the second contains a very comprehensive explanation of Statistics's invocation: expected input(s), output(s), and a full description of all the options.

All options and their values:: $ aststatistics -P
Inputs/Outputs and options:: $ info aststatistics
Full section in manual/book:: $ info Statistics
Full Gnuastro manual/book:: $ info gnuastro

If you couldn't find your answer in the manual, you can get direct help from experienced Gnuastro users and developers. For more information, please run:

$ info help-gnuastro

Statistics options:

Input:

-c, --column=STR

Column number (counting from 1) or search string.

--greaterequal2=FLT

Upper range of 2nd dim in 2D histograms.

-g, --greaterequal=FLT

Only use values greater-equal than this.

-h, --hdu=STR/INT

Extension name or number of input data.

-I, --ignorecase

Ignore case in matching/searching columns.

--lessthan2=FLT

Lower range of 2nd dim in 2D histograms.

-l, --lessthan=FLT

Only use values less than this.

-Q, --qrange=FLT[,FLT]

Quantile range: one (from Q to 1-Q) or two.

--searchin=STR

Select column(s): 'name', 'unit', 'comment'.

--stdintimeout=INT

Micro-seconds to wait for standard input.

Tessellation (tile grid):

--checktiles

Tile IDs in an image, the size of input.

-F, --remainderfrac=FLT

Fraction of remainder to split last tile.

--interpmetric=STR

Interpolation metric (radial, manhattan).

--interpnumngb=INT

No. of neighbors to use for interpolation.

--interponlyblank

Only interpolate over the blank tiles.

-i, --interpolate

Interpolate over blank tiles to fill them.

-M, --numchannels=INT[,..] No. of channels in dim.s (FITS order).

--oneelempertile

Display 1 element/tile, not full input res.

--workoverch

Work (not tile) over channel edges.

-Z, --tilesize=INT[,INT]

Regular tile size on dim.s (FITS order).

Output:

-D, --dontdelete

Don't delete output if it exists.

-K, --keepinputdir

Keep input directory for automatic output.

--outfitsnocommit

No Git commit in 0-th HDU of output FITS.

--outfitsnoconfig

No metadata in 0-th HDU of output FITS.

--outfitsnodate

No 'DATE' in 0-th HDU of output FITS.

--outfitsnoversions

No versions in 0-th HDU of output FITS.

-o, --output=STR

Output file name.

--tableformat=STR

Table fmt: 'fits-ascii', 'fits-binary', 'txt'.

--wcslinearmatrix=STR

WCS linear matrix of output ('pc' or 'cd').

Single value measurements

--concentration=FLT[,...]

Uniform dist: 1.0; higher: concentrated.

-d, --std

Standad deviation.

-E, --median

Median.

--mad

Median absolute deviation.

--madclip-mad

MAD-clipped Median Absolute Deviation.

--madclip-mean

MAD-clipped mean.

--madclip-median

MAD-clipped median.

--madclip-number

Number of elements after MAD-clipping.

--madclip-std

MAD-clipped standard deviation.

--maximum

Maximum.

--minimum

Minimum.

--modequant

Mode quantile (see --mode)

--modesym

Mode symmetricity (see --mode).

--modesymvalue

Value at mode symmetricity (see --mode).

-m, --mean

Mean.

--number

Number (non-blank).

-O, --mode

Mode (Appendix C of arXiv:1505.01664).

--quantfunc=FLT[,...]

Quantile function (multiple values acceptable).

--quantofmean

Quantile of the mean.

--sigclip-mad

Sigma-clipped Median Absolute Deviation.

--sigclip-mean

Sigma-clipped mean.

--sigclip-median

Sigma-clipped median.

--sigclip-number

Number of elements after sigma-clipping.

--sigclip-std

Sigma-clipped standard deviation.

--sum

Sum.

-u, --quantile=FLT[,...]

Quantile (multiple values acceptable).

Particular calculation

--asciicfp

Print an ASCII cumulative frequency plot.

-A, --asciihist

Print an ASCII histogram.

-C, --cumulative

Save the cumulative frequency plot in output.

--histogram2d=STR

2D histogram (as 'table' or 'image').

-H, --histogram

Save the histogram in output.

--madclip

Overall MAD-clipping (see '--mclipparams')

--mirror=FLT

Save the histogram and CFP of the mirror dist.

-R, --contour=STR

Contour levels, save in PGFPlots format.

-s, --sigmaclip

Overall sigma-clipping (see '--sclipparams')

-t, --ontile

Single values on separate tiles, not full input.

-y, --sky

Find the Sky and its STD over the tessellation.

Sky and Sky STD settings

-a, --outliernumngb=INT

Num neighboring tiles to look for outlier.

--checksky

Store steps in '_sky_steps.fits' file.

--checkskynointerp

Similar to --checksky, stops at interpolation.

--ignoreblankintiles

Don't write input's blanks in the tiled output.

--khdu=STR

HDU/extension name or number of kernel.

-k, --kernel=FITS

File name of kernel to convolve input.

--mclipparams=FLT,FLT

MAD clip: Multiple, and tolerance/number.

--meanmedqdiff=FLT

Max. mode and median quantile diff. per tile.

--mirrordist=FLT

Max. distance (error multip.) to find mode.

--outliersclip=FLT,FLT Sigma-clip params for qthresh outliers.

--outliersigma=FLT

Multiple of sigma to define outliers.

--sclipparams=FLT,FLT

Sigma clip: Multiple, and tolerance/number.

--smoothwidth=INT

Sky: flat kernel width to smooth interpolated.

Histogram and CFP settings

--asciiheight=INT

Height of ASCII histogram or CFP plots.

--manualbinrange

Set min/max of bins manually, not from data.

--maxbinone

Scale such that the maximum bin has value of one.

--numasciibins=INT

No. of bins in ASCII histogram or CFP plots.

--numbins=INT

No. of bins in histogram or CFP tables.

--numbins2=INT

No. of bins in second-dim of 2D histogram.

-n, --normalize

Set sum of all bins to 1.

--onebinstart=FLT

Shift bins so one bin starts on this value.

--onebinstart2=FLT

Similar to --onebinstart, for 2D histogram

Fitting (1 dimensional)

--fitestimate=STR/FLT

Estimate fitted func. (number, file or 'self').

--fitestimatecol=STR/INT

Column containing the --fitestimate values.

--fitestimatehdu=STR/INT

HDU containing the --fitestimate values.

--fitrobust=STR

The robust function name to use.

--fitweight=STR

Input weight: 'std', 'var' or 'inv-variance'.

-f, --fit=STR

Type of fitting: 'linear', 'linear-weighted', 'linear-no-constant', 'linear-no-constant-weighted', 'polynomial', 'polynomial-weighted', 'polynomial-robust'.

-p, --fitmaxpower=INT