pyblp.ProblemResults.extract_diagonal_means

ProblemResults.extract_diagonal_means(matrices, market_id=None)

Extract means of diagonals from stacked \(J_t \times J_t\) matrices for each market \(t\).

Parameters

• matrices (array-like) – Stacked matrices, such as estimates of \(\varepsilon\), computed by ProblemResults.compute_elasticities(); \(\mathscr{D}\), computed by ProblemResults.compute_diversion_ratios(); \(\bar{\mathscr{D}}\), computed by ProblemResults.compute_long_run_diversion_ratios(); or \(s_{ijt}\) computed by ProblemResults.compute_probabilities().

• market_id (object, optional) – ID of the market in which to extract diagonal means. By default, diagonal means are extracted in all markets and stacked.

Returns

Stacked diagonal means. If market_id was not specified, diagonal means are extracted in each market \(t\) and stacked. If the matrices are estimates of \(\varepsilon\), the mean of a diagonal is a market’s mean own elasticity of demand; if they are estimates of \(\mathscr{D}\) or \(\bar{\mathscr{D}}\), the mean of a diagonal is a market’s mean diversion ratio to the outside good. Rows are in the same order as Problem.unique_market_ids.

Return type

ndarray

Examples

• Tutorial