sportran.md.aic¶

Functions

`dct_AIC`(yk[, theory_var])	AIC[K] = sum_{k>K} c_k^2/theory_var + 2*(K+1) Assumiamo di tenere tutti i k <= K.
`dct_AICc`(yk[, theory_var])	AICc[K] = AIC[K] + 2(K+1)(K+2)/(NF-K-2) Assumiamo di tenere tutti i k <= K.
`dct_aic_ab`(yk, theory_var[, A, B])	AIC[K] = sum_{k>K} c_k^2/theory_var + 2*K Assumiamo di tenere tutti i k <= K.
`grid_statistics`(grid, density[, grid2])	Compute distribution mean and std. media = sum_i (density[i] * grid[i]) std = sqrt( sum_i (density[i] * grid[i]^2) - media^2 ) oppure = sqrt( sum_i (density[i] * grid2[i]) - media^2 ).
`produce_p`(aic[, method, force_normalize])
`produce_p_density`(p, sigma, mean[, grid, ...])

sportran.md.aic.dct_AIC(yk, theory_var=None)¶: AIC[K] = sum_{k>K} c_k^2/theory_var + 2*(K+1) Assumiamo di tenere tutti i k <= K.

sportran.md.aic.dct_AICc(yk, theory_var=None)¶: AICc[K] = AIC[K] + 2*(K+1)*(K+2)/(NF-K-2) Assumiamo di tenere tutti i k <= K.

sportran.md.aic.dct_aic_ab(yk, theory_var, A=1.0, B=2.0)¶: AIC[K] = sum_{k>K} c_k^2/theory_var + 2*K Assumiamo di tenere tutti i k <= K.

sportran.md.aic.grid_statistics(grid, density, grid2=None)¶: Compute distribution mean and std. media = sum_i (density[i] * grid[i]) std = sqrt( sum_i (density[i] * grid[i]^2) - media^2 )

oppure = sqrt( sum_i (density[i] * grid2[i]) - media^2 )

sportran.md.aic.produce_p_density(p, sigma, mean, grid=None, grid_size=1000)¶