# cline


Under this identifier are defined the parameters to compute the critical and the caustic lines.



## nplane int float [float]


`int` defines the number of Source Plane for which will be computed the critical and caustic lines. 
The list of `float` arguments give the redshift of these planes. 
Results are put in 2 different files: ”ce.dat” (external critic and caustic lines), ”ci.dat” (internal critic and caustic lines). ”ce.dat” and ”ci.dat” are 5-columns ASCII files with the format: ${j\, xIi\, yIi\, xSi\, ySi }$: $j$ is the line identifier (we can have more than one external or internal lines). $xIi$ and $yIi$ are the coordinates of the critical lines. $xSi$ and $ySi$ are the coordinates of the corresponding caustic lines.


## zonemult int1 int2 filename


`int1 = 0`: if false `int1 =1`: if true. If true and if `int` of [nplan]() equals 1, will determine for the redshift `float` of [nplan]() the image multiplicity of each pixel ( `int2`$\times$`int2` frame) of the image plane (area defined by `dmax` ). Results are written in the pixel-frame file `filename` (fits file) If [nplan]() not equals 1, zonemult is not executed, and a WARNING is displayed.



## dmax float


Defines the area (`xmin= -float` , `xmax = float` ; `ymin= -float` , `ymax = float` ) where the critical lines are search. `float` is expressed in arcseconds. A typical value of `dmax` is 30. Default: the value defined in `champ` section.



## algorithm string


string can be `MARCHINGSQUARE` (default) or `SNAKE`. 

Select one of the two algorithms for the computation of the critical and caustic lines. The snake algorithm is the original algorithm implemented in LensTool. For each of the first nlens crit clumps of the lens model, the algorithm starts from the centre of the clump and looks for a point on a surrounding critical line (locus of the space where amplification is infinite). Then, it tries to follow this critical line and to go back to its original starting point. The marching squares algorithm defines a first square with the `dmax` or the `champ` keywords. Then, it divides this first cube in 4 small squares. According to their size and their amplification values in the centre and the 4 corners, each square is divided or not in further small squares. If the field is rectangular, the greater value between the width and the height of the field is choosen as the square size. As long as the size of a square is greater than limitHigh, it is automatically divided in 4 small squares. The size of a square cannot be lower than limitLow. The marching squares algorithm is slower than the line-following snake algorithm but gives always the full contour of the critical lines. It is less sensitive to small irregularities in the contour. The snake algorithm always returns a connected contour. Default algorithm is marchingsquares.



## step float


For the line-following algorithm : Defines the step between each search i.e. it represents the typical distance in arcsecs between each point of the external critical lines. For internal critical lines half this value is taken. To improve the definition of the critic and caustic lines, use smaller values such as $0.5”$ or even $0.2”$. Default is $1.”$



## limitLow float

For the marching cube algorithm : Defines the smaller size of a square i.e. the size of a square is compared to this value to decide between dividing again in 4 squares or stopping the division. To improve the definition of the critic and caustic lines, use smaller values such as $0.5”$ or even $0.2”$. This implies more computation time. Default is $1”$.



## limitHigh float

For the marching cube algorithm : Defines the higher size of a square i.e. the size of a square cannot be higher than this value. A square with a size above this value is automatically divided in 4 squares. Decrease this value to remove holes in the critical lines and improve the detection of critical lines around isolated galaxies. This implies more computation time. Default is 10”.