|
CompoundPye
0.93
Modelling and Simulation Framework for Neural Networks of Arthropod Compound Eyes
|
Contains filter functions, creating input filters that will be multiplied with an array of input intensities that a sensor 'observes'. More...
Functions | |
| def | two_dim_gauss |
| Creates a two-dimensional gaussian filter-array. More... | |
| def | one_dim_gauss |
| Creates a one-dimensional gaussian filter-array. More... | |
Contains filter functions, creating input filters that will be multiplied with an array of input intensities that a sensor 'observes'.
Currently contains only one- and two-dimensional gaussian filter.
| def CompoundPye.src.Sensors.filter_funcs.one_dim_gauss | ( | cut_off, | |
| px, | |||
| rel_sigma, | |||
amplitude = 1, |
|||
search_step = 10 |
|||
| ) |
Creates a one-dimensional gaussian filter-array.
The size of the array is intrinsically defined by the parameter cut_off. cut_off is the lowest relative value that is still to be included in the filter-array. For example, if cut_off is 0.1, the filter-array stretches up to a value i, for which exp(-i^2/2/sigma^2)>0.1, while exp(-(i+1)^2/2/sigma^2)<=0.1.
| cut_off | Sets the relative value that is still to be included in the array; defines filter-array size intrinsically. |
| sigma | Defines width of the gaussian function.. |
| amplitude | Amplitude of the gaussian function at postition i=0. |
| search_step | One can increase or decrease computational efficiency by changing the search_step; never tested how well this works, though. |
| def CompoundPye.src.Sensors.filter_funcs.two_dim_gauss | ( | cut_off, | |
| px, | |||
| rel_sigma, | |||
amplitude = 1, |
|||
search_step = 10 |
|||
| ) |
Creates a two-dimensional gaussian filter-array.
The size of the array is intrinsically defined by the parameter cut_off. cut_off is the lowest relative value that is still to be included in the filter-array, looking at each direction individually. For example, if cut_off is 0.1, the filter-array stretches in x-direction up to a value x_i, for which exp(-x_i^2/2/sigma_x^2)>0.1, while exp(-x_(i+1)^2/2/sigma_x^2)<=0.1.
| cut_off | Sets the relative value that is still to be included in the array; defines filter-array size intrinsically. |
| sigma | List or array of two sigma values, one for each input dimension. |
| amplitude | Amplitude of the gaussian function at (x,y)=(0,0). |
| search_step | One can increase or decrease computational efficiency by changing the search_step; never tested how well this works, though. |