Using Config Files, Hidden Parameters and Plotting Behavior¶

Hidden Parameters¶

The most important parameters (Young’s modulus, Poisson’s ratio, gel height, window size and overlap for Particle Image Velocimetry) can be set in the clickpoints addons window. However, some parameters can only be set by using config files. Config files also allow you to set a wide range of plotting options, most notably: Color maps, the length of arrows, color bar positioning and dimensions and maximum and minimum Color maps, the length of arrows, color bar positioning and dimensions and maximum and minimum displayed values for all frames. You can find a complete list of parameters in Overview of Analysis Parameters and Overview of Plotting Parameters.

Using Config Files¶

You can provide a config file by placing a file “config.yaml” in the same folder as your database (.cdb) file. YAML is a standard format for configuration files. In YAML files parameters are defined by a series of indented key words like this:

key1:
  key2:
    key3: value

You can find a config.yaml file that reproduces all default parameters in the example data. The config file distinguishes two classes of parameters at the first level: The “analysis_parameters” and the “fig_parameters”. “analysis_parameters” set all parameters that are used by the program not related to plotting. They can only be numbers or strings. For example to change the minimum object size and change the initial value for the Poisson’s ration, your config file should contain this:

analysis_parameters: # setting parameters for the analysis
  sigma: 0.4 # changing the Poisson's ratio set at the script start up
  min_obj_size: 1500 # changing the minimal object size (cannot be changed in the addon window)

“fig_parameters” control the plotting. Most parameters can be specified as a single value that will apply to all plots. Alternatively you can specify a second key that defines a type of plot first and then a value. This value will only apply to the plot type you specified. All other plot types will use a default value. Possible plot types are:

“deformation”    - Quiver plot of the deformation field
“traction” - Quiver plot of the traction forces
“FEM borders” - Plot of the line tension along cell-cell borders
“stress map”  -  Plot of the mean normal stress
“energy_points”. - Plot of the strain energy density

Let’s say you want to

A): Change the color map to “jet” (all matplotlib color maps are accepted) for all plots.
B): Set the maximum value of the colormap only in the deformation plots to 10 pixels.
C): Generate a map of strain energy, when analysing cell colonies.

Then your yaml file should look like this:

fig_parameters:   # setting plotting parameters

  cmap: "jet" # changing the colormap for the deformation plot

  vmin: None  # minimal value displayed in the colormap, none means an automatic value for each frame
          # this applies to all plot types if none is specified
  vmax:       # maximum value displayed in the colormap
    deformation: 10 # applies only to deformation plots. All other plots use a default value (None)

 plots:  # defining which plots are produced
     colony: # you need to specify whether this is for "colony" or "cell layer"
      - "deformation" # this needs a list of values (marked by "-")
      - "traction"
      - "FEM_borders"
      - "stress_map"
      - "energy_points"

Hint

It is best to mark strings with quotation marks (“”). You can use None, True or False without quotation marks to set None or boolean values.

Overview of Analysis Parameters¶

Parameter	Default Value	Type	Significance
Main Parameters
sigma	0.49	int,float	Poisson’s ration of the substrate.
young	49000	int,float	Young’s modulus of the substrate in Pa.
pixelsize	0.201	int,float	Pixel size of the images of the beads in µm.
window_size	20	int,float	Size of the windows for PIV (Particle Image Velocimetry) in µm.
overlap	19	int,float	Size of the overlap for PIV in µm.
FEM_mode	“colony”	string	Analyzing colonies or cell layer. This changes the behavior, concerning which masks are used, which plots are generated and what area is used for stress measurements.
Hidden Parameters
std_factor	15	int,float	Additional filter for the deformation field. Deformations greater then (\(µ\) and \(\sigma\): mean and standard deviation of the norm of deformations) are replaced by the local mean deformation.
edge_padding	0.1	float	All masks are cut of close to the image edge, i.e. if they are closer then edge_padding*axis_length. For FEM analysis, all pixels at this edge are fixed so that no displacement perpendicular to the axis is allowed.
padding_cell_layer	0.2	float	If you are analyzing cell layers, and additional region close to the image edge is ignored when analyzing stresses, to avoid boundary effects. The effectively ignored region for cell layers is edge_padding + padding_cell_layer.
sigma_cells	0.5	float	Poisson’s ratio of the cell sheet in the MSM algortihm. This parameter should have negligible influence on the resulting stresses.
min_obj_size	1500	int	Minimum size of an object (cell or cell colony). All masks are added up and all encircled areas are filled to determine the object size.
cv_pad	0	int,float	File names. Include the ending (e.g. “.png”)
TFM_mode	“finite_thickness”	string	Using a TFM algorithm assuming either finite substrate thickness (“finite_thickness”) for infinite substrate thickness (“infinte_thickness”). Always use “finite_thickness”.

Overview of Plotting Parameters¶

Parameter	Default Value	Type	Significance
file_names	specific	string	File names. Include the ending (e.g. “.png”)
cmap	“rainbow”	string	Color maps. All matplotlib color maps are accepted.
vmin	None	float, int, None	Minimal value of the color bar. None for automatic selection.
vmax	None	float, int, None	Maximal value of the color bar. None for automatic selection.
Color bar Parameters
cbar_style	“clickpoints”	“clickpoints” or “outside”	Specifies whether the color bar is plotted inside or outside of the image. Plotting the color bar outside will lead to misaligned images in clickpoints.
cbar_axes_fraction	0.2	float < 1	Height of the color bar when using cbar_style “outside”. This number signifies the fraction of the length of the original image axis.
cbar_width	“2%”	string	Width of the color bar when using cbar_style “clickpoints”. Has to be a string signifying the percentage of of the original image axis.
cbar_height	“50%”	string	Height of the color bar when using cbar_style “clickpoints”. Has to be a string signifying the percentage of of the original image axis.
cbar_borderpad	6	int	Distance between the color bar and the right image edge.
cbar_str	specific	string	Title of the color bar. Use quotation marks (“”) in the config file.
cbar_title_pad	10	int	Distance between the color bar and the color bar title.
cbar_tick_label_size	15	int	Size of the color bar tick labels.
Arrows in Deformation and Traction Fields
filter_factor	1	float,int > 0	Factor that defines how many arrows are filtered out for plotting (traction and deformation fields). A high filter_factor means less arrows are plotted.
scale_ratio	0.2	float (0,1]	Length of the arrows (deformation and traction fields). Arrows are scaled so that the longest arrow has the length scale_ratio * longest image axis.
width	0.002	float	Width of the arrow shaft (traction and deformation fields).
headlength	3	float,int	Length of the arrow heads (traction and deformation fields).
headwidth	3	float,int	Width of the arrow head (traction and deformation fields)

+++

Plotting the Line Tensions
background_color	“#330033”	string, tuple	Color of the background. Can be any color format accepted by matplotlib. You can use “cmap_0” to use the color of zero in the colormap used for the plot.
plot_t_vecs	False	bool	Plotting the line tension vectors.
plot_n_arrows	False	bool	Plotting the normal vectors of the cell boundary lines.
linewidth	4	int, float	Width of the lines representing the cell boundary lines.
border_arrow_filter	1	int	Filter defining how many arrows are plotted along the cell boundary lines. Only every n-th arrow is plotted, where n is the border_arrow_filter.
boundary_resolution	6	int	Smoothness of the lines representing the cell boundary lines. A high boundary_resolution means less smooth plotting. means less smooth plotting. Very high values will cost a considerable amount of computation time.
Choosing which Plots are generated
plots - colony	“deformation”, “traction”, “FEM_borders”, “stress map”		List of plots that are produced in “colony” or “cell layer” mode.
plots - cell layer	“deformation”, “traction”, “FEM_borders”, “stress map”, “energy points”		List of plots that are produced in “colony” or “cell layer” mode.