next up previous contents
Next: Intrinsics and Technicalities Up: Some Useful Tricks with Previous: Other Fields   Contents

Exercises

  1. In the Purkinje cell tutorial, plot the overall contribution of one of the $ \mathrm{Ca}^{2+}$ channels in a graph. Add code that allows to inspect the overall $ \mathrm{Ca}^{2+}$ influx in the cell (the $ \mathrm{Ca}^{2+}$ currents are modeled with two separate channels).
  2. To randomize the properties of a population before a simulation, I once encountered code a bit like the following (I summarized it to the pieces of interest):
      echo "Randomizing granule cells"
      for (i = {number_granule_cells}; i > 0; i = i - 1)
        pushe /granule_cell_layer/Granule[{i-1}]/soma
        initvm = {rand {Vm_init_lb} {Vm_init_ub}}
        setfield . initVm {initvm}
        setfield . Vm     {initvm}
        setfield . Em     {rand {Granule_E_leak_lb} {Granule_E_leak_ub}}
        call /granule_cell_layer/Granule[{i-1}]/solve \
          HPUT /granule_cell_layer/Granule[{i-1}]/soma
        pope
        pushe /granule_cell_layer/Granule[{i-1}]/soma/mf_AMPA
        setfield . \
                gmax {{getfield . gmax}
                      * (1 + {weight_distribution} * {rand -1 1})}
        call /granule_cell_layer/Granule[{i-1}]/solve HPUT .
        pope
      end
    
    The man that sent me the code complained that hsolve is indeed very fast, but the setup of hsolve takes ages. What is redundant in the code above? Why? Is this causing a bottleneck during the simulation?



2002-11-15