Automatic response file generationΒΆ

This allows the user to make requests such as: Write responses for all pre-synaptic inputs to a given cell that have weight greater than 0.1. Or: Write responses for all units that have orientation preferences between 0 and 40 degrees, within a specific population.

The is acheived by running ‘wm’ twice. First, to generate a response (.rsp) file, and then to run with the usual ‘mod’ option using the newly generated response file. The response file generation step is executed with a command of this form:

wm rsp_gen  [your.moo]  [your.rsg]  [your.rsp]

where the first argument is a model parameter (.moo) file, the second is a response generation (.rsg) file, and the third is the output response (.rsp) file to be generated. The .rsg file should look exactly like any .rsp file, except the following commands can be used, which will be expanded into simple response request lines:

gen_save_pop_lay_to_unit_as_XXX [data_type] [sampling] [pop_pre]
                                [pop_post] [x] [y] [z] [minw] [data_ID]
where,
  • XXX - name to be used as the record name in the .nd file, typically the pre-syn population name is chosen.
  • data_type - “s” or “f”
  • sampling - given in samples/s
  • pop_pre - pre-synaptic population name
  • pop_post - post-synaptic population name
  • x,y,z - coordinates of unit within population
  • minw - minimum weight to consider (-1.0 will allow all, if all w >= 0)
  • data_ID - a particular identifier that is associated with a signal within the model.

Save layer-to-layer connections:

gen_save_pop_lay_to_lay_as_XXX [data_type] [sampling] [pop_pre]
                               [pop_post] [minw] [data_ID]
where,
  • XXX - name to be used as the record name in the .nd file, typically the pre-syn population name is chosen.
  • data_type - “s” or “f”
  • sampling - given in samples/s
  • pop_pre - pre-synaptic population name
  • pop_post - post-synaptic population name
  • minw - minimum weight to consider (-1.0 will allow all, if all w >= 0)
  • data_ID - a particular identifier that is associated with a signal within the model.

NOTE: Units marked for saving will not be repeated again in the .rsp file for a later instance of the above gen_save_ commands.:

gen_save_pop_multi_lay_as_XXX [data_type] [sampling] [pop_pre]
                              [...] [pop_post] [optype] [fcrit] [data_ID]
where,
  • XXX - name to be used as the record name in the .nd file, typically the pre-syn population name is chosen.
  • data_type - “s” or “f”
  • sampling - given in samples/s
  • pop_pre - pre-synaptic population name
  • [...] - additional populations that must be included in this order
  • pop_post - post-synaptic population name
  • optype - condition operation, e.g., “w_agg_min” to set a minimum aggregate weight
  • fcrit - critical value, allow weights > fcrit (0 will allow all positive weights)
  • data_ID - a particular identifier that is associated with a signal within the model.

Save multiple layer connections:

gen_save_pop_mlay_unit_as_XXX [data_type] [sampling] [pop_pre]
                              [...] [pop_post] [x_post] [y_post] [z_post]
                              [optype] [fcrit] [data_ID]
where,
  • XXX - name to be used as the record name in the .nd file, typically the pre-syn population name is chosen.
  • data_type - “s” or “f”
  • sampling - given in samples/s
  • pop_pre - pre-synaptic population name
  • [...] - additional populations that must be included in this order
  • pop_post - post-synaptic population name
  • x,y,z_post - coords for post-synaptic unit
  • optype - condition operation, e.g., “w_agg_min” to set a minimum aggregate weight
  • fcrit - critical value, allow weights > fcrit (0 will allow all positive weights)
  • data_ID - a particular identifier that is associated with a signal within the model.

Save depending on orientation:

gen_save_pop_lay_ori_as_XXX [data_type] [sampling] [pop]
                            [op_type] [val1] [val2] [data_ID]
where,
  • XXX - name to be used as the record name in the .nd file, typically the pre-syn population name is chosen.
  • data_type - “s” or “f”
  • sampling - given in samples/s
  • pop - population name
  • op_type - “dir180_ctr_dev”
  • val1 - the central value for the range of ori’s to accept
  • val2 - the maximum deviation away from ‘val1’
  • data_ID - a particular identifier that is associated with a signal within the model.

This will include all units in the specified population that have values for the ‘ori’ attribute that are no further away from ‘val1’ than ‘val2’.