Backend selection#

get_backend_info(backend=None)#

Return backend metadata without creating a network.

Parameters:: backend -- Backend name. Use None to select an available backend.
Returns:: The backend name, library path, and sparse solver name.

class GroupCase(name, rows, skipped=<factory>)#

Store member results for one case in a group study.

static row_json(r, header)#: One row of the SQL deliverability table: the member's weight and request, the group's answer, the member's share of it, the full-versus-partial verdict, and what bound the group.

static upgrade_id(r)#: This study's identity columns on a row of the upgrades table. The members share one transfer, so each drives the element in its own proportion and the shares add back up to the group's.

static cols(header)#: Return report columns for the group-deliverability table.

static binding_cols(header)#: The binding-case table of a case set, in this study's columns.

class GroupMember(kind, weight, request_mw, group_mw, headroom_mw, *, bus, label, case, dc, ac=None, limit_type='unconstrained', limiter=None)#

Store one member's weighted share of group delivery.

headroom_mw = group_mw * weight. request_mw uses the same case dispatch assumption that produced the weight. Every member shares the group's DC, optional AC, and limiter results.

property verdict#: Whether the group delivers this member's whole request here.

class CaseSetReport(header, cases, binding, stats=<factory>, timing=<factory>)#

Store interconnection-screen results for a set of cases.

cases holds the per-case records in declaration order and is never empty. binding holds each candidate's lowest-headroom case. timing holds the run's seconds and counts per tier; the screen runs every case in one call, so the clock belongs to the set.

classmethod of(header, cases, *, timing=None, **stats)#: Build a report and derive binding rows from its case results.

doc()#: Return the result document: the scalar header, with the column bindings for the row-tables, which live on the queryable slot and are queried in SQL.

as_tables()#

Row-tables for the SQL query surface: binding (one row per candidate, its worst case), the per-case table named by the screen that ran, screen, headroom or deliverability, holding every candidate in every case, and upgrades.

upgrades is that same per-case set exploded over each candidate's ranked NETWORK limiters (result, never the capped answer), carrying the flow the candidate drives on each element. Grouping it by element is proportional-impact cost allocation over a result that already exists. This requires one query and no second solve.

column_quantities()#

Return the quantity each projected column binds, keyed by table then column.

The screen that ran owns the row schema, so it owns the bindings too: the per-case table and the binding table hold the same columns, and the upgrade table holds that screen's identity columns beside what driver_share() writes.

to_markdown(path=None)#: Return the case-set report as Markdown.

class ScreenedCase(name, rows, skipped=<factory>)#

Store one case's rows and dropped contingency declarations.

The case receives screen-wide header data as an argument. A report pose supplies row serialization, upgrade identity, and column specifications.

classmethod binding_columns()#: Return the column bindings of the binding table.

classmethod upgrade_columns()#: Return the column bindings of the upgrades table.

classmethod binding_json(r, header)#: Return one row of the binding table.

classmethod table_rows(r, header)#: Return the per-case table rows one screened row projects.

classmethod upgrade_rows(r, header)#: Return one upgrades row per pre-existing violation and ranked network limiter.

rendered_rows(header)#: Return the rows used by this case's Markdown table.

classmethod rendered_binding(rows, header)#: Return the rows used by the binding Markdown table.

section(header)#: Return this case's table and footnotes without a heading.

class ScreenHeader(sink, contingencies, ranked, contingency_rating='normal', strength=False, scr_threshold=3.0)#

Describe the common settings for one interconnection screen.

sink identifies the withdrawal point. contingencies is the declaration the screen evaluated, as written. ranked identifies the result tier used to compare candidates. contingency_rating identifies the rating set applied after outages. scr_threshold defines a weak grid.

weak(scr)#: Return the weak-grid flag, or None when SCR is unavailable.