vivainsights.create_radar¶

create_radar: Parameterized radar-chart workflow (calc + viz + wrapper), in the same spirit as create_bar.

Core design¶

General-purpose: works with any HR attribute column.
Calculation pipeline:
- person-level aggregation within each group
- group-level aggregation
- minimum group size (mingroup)
- indexing modes: “total”, “none”, “ref_group”, “minmax”
Returns either a plot or a table.

Typical usage¶

>>> import vivainsights as vi
>>> from vivainsights.create_radar import create_radar
>>>
>>> pq_data = vi.load_pq_data()
>>> fig = create_radar(
...     data=pq_data,
...     metrics=[
...         "Copilot_actions_taken_in_Teams",
...         "Collaboration_hours",
...         "After_hours_collaboration_hours",
...         "Internal_network_size",
...     ],
...     hrvar="Organization",
... )

Return the indexed table instead of a plot:

>>> tbl = create_radar(
...     data=pq_data,
...     metrics=["Collaboration_hours", "Meetings_count"],
...     hrvar="Organization",
...     return_type="table",
... )

Reference a specific group as 100 (ref_group indexing):

>>> fig = create_radar(
...     data=pq_data,
...     metrics=["Collaboration_hours", "Meetings_count"],
...     hrvar="Organization",
...     index_mode="ref_group",
...     index_ref_group="Contoso Ltd",
... )

Min-max scaling to [0,100] within observed group ranges:

>>> fig = create_radar(
...     data=pq_data,
...     metrics=["Collaboration_hours", "Meetings_count", "Focus_hours"],
...     hrvar="Organization",
...     index_mode="minmax",
... )

vivainsights.create_radar.create_radar_calc(data, metrics, hrvar, id_col='PersonId', mingroup=5, agg='mean', index_mode='total', index_ref_group=None, dropna=True)[source]¶

Name¶

create_radar_calc

Description¶

Compute group-level metric values and (optionally) index them for radar plotting.

Steps:

Aggregate to person-level within each group (mean/median).
Aggregate the person-level values to the group level.
Enforce a minimum person count per group (mingroup).
Apply an indexing mode to make metrics comparable.

param data:

Standard Person Query data frame containing metrics, hrvar, and id_col.

type data:

pd.DataFrame

param metrics:

Numeric metric column names to summarise and index for the radar chart.

type metrics:

List[str]

param hrvar:

HR attribute column identifying the group for each person (e.g., “Organization”, “LevelDesignation”).

type hrvar:

str

param id_col:

Column uniquely identifying people for person-level aggregation.

type id_col:

str, default “PersonId”

param mingroup:

Minimum number of unique people required in a group to retain it.

type mingroup:

int, default 5

param agg:

Aggregation function for both person-level and group-level summaries.

type agg:

{“mean”,”median”}, default “mean”

param index_mode:

“total”: index each metric vs. the overall person-level average (Total = 100).
“ref_group”: index vs. a specific group given by index_ref_group (Ref = 100).
“minmax”: scale to [0,100] within the min-max of observed group values (per metric).
“none”: return raw (unindexed) group values.

type index_mode:

{“total”,”none”,”ref_group”,”minmax”}, default “total”

param index_ref_group:

Required when index_mode=”ref_group”. Name of the group to serve as reference (=100).

type index_ref_group:

Optional[str], default None

param dropna:

If True, drop rows with NA in any of [id_col, hrvar] + metrics prior to aggregation.

type dropna:

bool, default True

returns:

(group_level_indexed, ref) –

group_level_indexed: One row per group, wide across metrics. Values are indexed/scaled as per index_mode.
ref: The reference used for indexing: - For “total” / “ref_group”: a pd.Series of reference means/medians. - For “minmax”: a two-column DataFrame with per-metric min and max. - For “none”: empty Series.

rtype:

Tuple[pd.DataFrame, pd.Series]

vivainsights.create_radar.create_radar_viz(data, metrics, hrvar, fill_missing='zero', figsize=(8, 6), title=None, subtitle=None, caption=None)[source]¶

Name¶

create_radar_viz

Description¶

Render a radar (spider) chart from a wide, group-level table produced by create_radar_calc. Each row in data is plotted as a polygon across the supplied metrics in the given order.

param data:: One row per group, columns include hrvar and each of metrics. Values should already be indexed/scaled to comparable units (i.e. the output of create_radar_calc).
type data:: pd.DataFrame
param metrics:: Ordered list of metric columns to plot around the radar.
type metrics:: List[str]
param hrvar:: Column containing the group labels used in the legend.
type hrvar:: str
param fill_missing:: How to handle NA values in data before plotting: - “zero”: replace NA with 0 so polygons close correctly. - “none”: leave NA as-is (polygon may not render for that group).
type fill_missing:: str, default “zero”
param figsize:: Matplotlib figure size in inches (width, height).
type figsize:: Tuple[float, float], default (8, 6)
param title:: Top title for the figure.
type title:: Optional[str], default None
param subtitle:: Optional smaller line beneath the title (figure-level, not axes).
type subtitle:: Optional[str], default None
param caption:: Small text near the bottom of the figure (e.g., date range).
type caption:: Optional[str], default None
returns:: fig – The constructed matplotlib Figure.
rtype:: matplotlib.figure.Figure

vivainsights.create_radar.create_radar(data, metrics, hrvar='Organization', id_col='PersonId', mingroup=5, agg='mean', index_mode='total', index_ref_group=None, dropna=False, return_type='plot', figsize=(8, 6), title=None, subtitle=None, caption=None)[source]¶

Name¶

create_radar

Description¶

High-level convenience wrapper to compute group-level metrics and either:

return the indexed table (return_type=”table”), or
render a radar chart (return_type=”plot”).

param data:

Standard Person Query data frame containing at least metrics, id_col, and hrvar.

type data:

pd.DataFrame

param metrics:

Numeric metric columns to visualise (order determines the radar axes).

type metrics:

List[str]

param hrvar:

HR attribute column used for grouping (e.g., “Organization”, “LevelDesignation”).

type hrvar:

str, default “Organization”

param id_col:

Unique person identifier for person-level aggregation.

type id_col:

str, default “PersonId”

param mingroup:

Minimum unique person count per group.

type mingroup:

int, default 5

param agg:

Aggregation function for person- and group-level summaries.

type agg:

{“mean”,”median”}, default “mean”

param index_mode:

Indexing/scaling mode applied to group values prior to plotting.

type index_mode:

{“total”,”none”,”ref_group”,”minmax”}, default “total”

param index_ref_group:

Required when index_mode=”ref_group”. The name of the group that will be fixed at 100.

type index_ref_group:

Optional[str], default None

param dropna:

Drop rows with NA in required columns prior to aggregation.

type dropna:

bool, default False

param return_type:

“plot”: return a matplotlib Figure.
“table”: return the indexed group-level DataFrame.

type return_type:

{“plot”,”table”}, default “plot”

param figsize:

Figure size for the plot (ignored when return_type=”table”).

type figsize:

Tuple[float, float], default (8, 6)

param title:

Plot title. If None, a default title is inferred based on index_mode.

type title:

Optional[str], default None

param subtitle:

Optional subtitle line.

type subtitle:

Optional[str], default None

param caption:

Additional caption text appended after the auto-generated date range and index label, e.g. “caption” → “Data from … | Index: … | caption”. If None, only the date range and index label are shown.

type caption:

Optional[str], default None

returns:

If return_type=”plot”: a Figure containing the radar chart.
If return_type=”table”: the group-level indexed DataFrame.

rtype:

matplotlib.figure.Figure or pd.DataFrame