10 How far do people travel to Hubbard Hall?

The donor-geography chapter asked where the money comes from; this one asks where the people come from. Counts are registrations by the registering household’s distance band (straight-line miles — see Methodology); registrations without a geocodable address (about 2%) are dropped.

Code

import pandas as pd
import plotly.express as px
from hh.io import read_parquet
from hh.viz import style, pct
from hh.analytics.timing import WEEKDAY_ORDER

BANDS = ["[0, 1)", "[1, 5)", "[5, 10)", "[10, 20)", "[20, ∞)"]
BAND_COLORS = dict(zip(BANDS, ["#b7d3f6", "#6da7ec", "#3987e5", "#184f95", "#0d366b"]))

regs = read_parquet("processed", "registrations_enriched.parquet")
cat_order = ["theater", "music", "dance", "opera", "class"]
named = regs[regs["event_minorcat"].isin(cat_order)].assign(
    cat=lambda d: d["event_minorcat"]
)
combo = pd.concat([named, regs[regs["event_majorcat"] == "class"].assign(cat="class")])
combo = combo.dropna(subset=["distance_band"])

band_cat = (
    combo.groupby(["cat", "distance_band"], observed=True)
    .size()
    .unstack("distance_band")
    .reindex(index=cat_order, columns=BANDS)
    .fillna(0)
    .astype(int)
)
w10 = (band_cat[BANDS[:3]].sum(axis=1) / band_cat.sum(axis=1))
far = band_cat[BANDS[-1]] / band_cat.sum(axis=1)

Registrations by distance band and programming type:

Code

t = band_cat.copy()
t["within 10 mi"] = w10
t.style.format("{:,}", subset=BANDS).format({"within 10 mi": pct})

distance_band	[0, 1)	[1, 5)	[5, 10)	[10, 20)	[20, ∞)	within 10 mi
cat
theater	1175	1243	1255	1016	2001	54.9%
music	619	743	679	555	836	59.5%
dance	62	66	106	38	35	76.2%
opera	233	273	256	202	502	52.0%
class	3393	3252	3597	1851	1418	75.8%

Code

share = (
    band_cat.div(band_cat.sum(axis=1), axis=0)
    .reset_index()
    .melt(id_vars="cat", var_name="distance_band", value_name="share")
)
fig = px.bar(
    share, x="cat", y="share", color="distance_band",
    color_discrete_map=BAND_COLORS,
    category_orders={"cat": cat_order, "distance_band": BANDS},
)
fig.update_layout(
    xaxis_title="", yaxis_title="Share of registrations", legend_title="Distance band"
)
fig.update_yaxes(tickformat=".0%")
fig.update_traces(hovertemplate="%{y:.0%}<extra>%{fullData.name}</extra>")
style(fig).show()

Classes serve the immediate community — 76% of class registrations come from within 10 miles (dance is similar at 76%). Performances pull from a much wider region: only 55% of theater, 59% of music, and 52% of opera registrations are that local, and roughly a third of theater (30%) and opera (34%) registrations come from beyond 20 miles.

10.1 Do weekend theater audiences travel differently?

For theater only, registrations from each distance band on each day of the week — “do Friday/Saturday audiences travel farther than Sunday’s?”

Code

theater = regs[regs["event_minorcat"] == "theater"].dropna(subset=["weekday", "distance_band"])
day_band = (
    theater.groupby(["weekday", "distance_band"], observed=True)
    .size()
    .unstack("distance_band")
    .reindex(index=WEEKDAY_ORDER, columns=BANDS)
    .fillna(0)
    .astype(int)
)
far_day = (day_band[BANDS[-1]] / day_band.sum(axis=1))
t = day_band.copy()
t["beyond 20 mi"] = far_day
t.style.format("{:,}", subset=BANDS).format({"beyond 20 mi": pct})

distance_band	[0, 1)	[1, 5)	[5, 10)	[10, 20)	[20, ∞)	beyond 20 mi
weekday
Monday	0	0	0	0	0	nan%
Tuesday	3	6	6	4	9	32.1%
Wednesday	6	5	5	1	2	10.5%
Thursday	7	16	10	7	7	14.9%
Friday	332	412	342	250	434	24.5%
Saturday	555	533	557	425	851	29.1%
Sunday	272	271	335	329	698	36.6%

Sunday draws the farthest-traveling theater audience: 37% of Sunday registrations come from beyond 20 miles, versus 25% on Friday and 29% on Saturday. Friday is the most local night. That fits the within-run finding in When do people come to Hubbard Hall?: the Sunday matinee is not a neighborhood habit — it is when out-of-area visitors come.