Skip to main content

DataTable: grouping & pivot

Grouping, aggregation, and pivot are QCustomDataTablePro features (the commercial add-on — see Licensing).

from custom_widgets_pro import QCustomDataTablePro
from Custom_Widgets.QCustomDataTable import DataTableColumn

rows = [
    {"region": "East", "category": "Hardware", "amount": 120, "qty": 3},
    {"region": "East", "category": "Software", "amount": 300, "qty": 1},
    {"region": "West", "category": "Hardware", "amount": 90,  "qty": 2},
    {"region": "West", "category": "Software", "amount": 200, "qty": 4},
]

table = QCustomDataTablePro()
table.setColumns([DataTableColumn("region"), DataTableColumn("category"),
                  DataTableColumn("amount", type="number"),
                  DataTableColumn("qty", type="number")])
table.setData(rows)
table.show()

Grouping with aggregates​

Group by one or more column keys; the second argument maps a column to an aggregate shown on each group header:

table.groupBy(["region"], {"amount": "sum", "qty": "avg"})

Aggregates: sum, avg, count, min, max, first, last, or a callable fn(values) -> result. Group by several keys for nested groups:

table.groupBy(["region", "category"], {"amount": "sum"})

Expand / collapse​

Click a group header to toggle it, or drive it in code:

table.expandAllGroups()
table.collapseAllGroups()
table.groupToggled.connect(lambda path, expanded: print(path, "expanded" if expanded else "collapsed"))

Clear grouping to return to the flat, virtualized view:

table.clearGrouping()
print(table.isGrouped(), table.groupKeys())

Pivot (cross-tab)​

Reshape into a matrix: one row per index value, one column per distinct columns value, each cell an aggregate of values:

table.pivot(index="region", columns="category", values="amount", aggfunc="sum")
# -> columns: Region | Hardware | Software | Total

Options: pivot(index, columns, values, aggfunc="sum", totals=True, fill=None). index may be a list of keys for nested row headers. Restore the flat table with:

table.clearPivot()
print(table.isPivoted())

Grouping and pivot are mutually exclusive — starting one clears the other.