simrun ❭ synaptic_strength_fitting ❭ calculate_optimal_g

calculate_optimal_g¶

simrun.synaptic_strength_fitting.calculate_optimal_g(pdf)¶

Calculate the optimal synaptic conductance such that the EPSP statistics match empirical data.

This function calculates the optimal synaptic conductance by matching empirically observed EPSP statistics (mean, median, maximum) to a linear model. Each statistic provides a different estimate of the optimal g. The final optimal g is then a weighed average of the three statistics, where the weights for mean:median:max are 2:2:1 respectively.

This function is used in PSPs to calculate the optimal synaptic conductance for each celltype.

Parameters:¶

pdf (pd.DataFrame) – A table containing the empirical EPSP statistics (mean, median and maximum), and linear fits for each statistic.

Returns:¶

None. Updates the original table inplace. Adds the columns optimal g, optimal g mean, optimal g median, and optimal g max.

See also

linear_fit_pdf() for the linear model that relates the EPSP statistics to the synaptic conductance.

Example:

>>> pdf = psp.get_summary_statistics(method='dynamic_baseline')
>>> fit = linear_fit_pdf(pdf)
>>> fit.head()
    EPSP mean_offset  EPSP mean_slope  EPSP med_offset  EPSP med_slope  EPSP max_offset  EPSP max_slope EPSP_std_offset EPSP_std_slope
...
>>> measured_data
            EPSP_mean_measured  EPSP_med_measured   EPSP_max_measured
celltype_1  ...
celltype_2  ...
>>> pdf = pd.concat([fit, measured_data], axis=1)
>>> calculate_optimal_g(pdf)
>>> pdf[cell_type 1']['optimal g']
1.85