In the older pytests for STDP synapses (e.g. test_stdp_pl_synapse_hom.py, test_stdp_synapse.py) hard-coded spike times are fed to alpha psc neurons, whose spike output is then used for a manual reproduction of the weight drift & subsequently compared. The issue is that, due to the nature of the alpha neurons and the high weights set from generators to neuron populations, many of the hardcoded times (and therefore potential corner cases for testing) end up ignored.
In the older pytests for STDP synapses (e.g. test_stdp_pl_synapse_hom.py, test_stdp_synapse.py) hard-coded spike times are fed to alpha psc neurons, whose spike output is then used for a manual reproduction of the weight drift & subsequently compared. The issue is that, due to the nature of the alpha neurons and the high weights set from generators to neuron populations, many of the hardcoded times (and therefore potential corner cases for testing) end up ignored.