Google search volumes for 2024 solar eclipse

Input file downloaded manually from https://trends.google.com/ because of API rate limits.

import pandas as pd

df_eclipse_2024 = pd.read_csv('geoMap_eclipse_2024.csv',skiprows=0)
df_eclipse_2024

	Zone	volume
0	Burlington VT-Plattsburgh NY	100
1	Presque Isle ME	96
2	Paducah KY-Cape Girardeau MO-Harrisburg-Mount ...	88
3	Bangor ME	85
4	Portland-Auburn ME	84
...	...	...
205	Seattle-Tacoma WA	27
206	Anchorage AK	26
207	Honolulu HI	25
208	Juneau AK	23
209	Fairbanks AK	22

210 rows × 2 columns

Now we need to find the geographical area corresponding to every "Zone". Google Trends uses zones (Designated Market Areas = DMAs) that do not follow exactly US counties or particular administrative divisions. Somewhere through Kaggle I found a list of all US counties, with the DMA they belong to. I had to manually edit it because it switches Portland (Oregon) and Portland (Maine).

google_dma = pd.read_csv('GoogleTrends_CountyDMA_Mapping.csv')
google_dma

	STATE	COUNTY	STATEFP	CNTYFP	CNTYTVHH	DMAINDEX	DMA	GOOGLE_DMA
0	AK	Anchorage Borough	2	20	95010	156	ANCHORAGE (AK) ...	Anchorage AK
1	AK	Kenai Peninsula Borough	2	122	17600	156	ANCHORAGE (AK) ...	Anchorage AK
2	AK	Matanuska-Susitna Borough	2	170	20130	156	ANCHORAGE (AK) ...	Anchorage AK
3	AK	Fairbanks North Star Borough	2	90	31860	203	FAIRBANKS (AK) ...	Fairbanks AK
4	AK	Juneau Borough	2	110	25270	206	JUNEAU (AK) ...	Juneau AK
...	...	...	...	...	...	...	...	...
3115	WY	Weston	56	45	2460	175	RAPID CITY (SD) ...	Rapid City SD
3116	WY	Lincoln	56	23	5240	36	SALT LAKE CITY (UT) ...	Salt Lake City UT
3117	WY	Sublette	56	35	2260	36	SALT LAKE CITY (UT) ...	Salt Lake City UT
3118	WY	Sweetwater	56	37	13530	36	SALT LAKE CITY (UT) ...	Salt Lake City UT
3119	WY	Uinta	56	41	6800	36	SALT LAKE CITY (UT) ...	Salt Lake City UT

3120 rows × 8 columns

Every US county has a 5-digit FIPS code, whose first two digits represent the US state. In the above table, the numbers corresponding to the state and to the county are given in separate columns, so we merge them:

fips = [ str(google_dma['STATEFP'][iii]).zfill(2) + str(google_dma['CNTYFP'][iii]).zfill(3) for iii in range(len(google_dma)) ]
google_dma['fips'] = fips
google_dma

	STATE	COUNTY	STATEFP	CNTYFP	CNTYTVHH	DMAINDEX	DMA	GOOGLE_DMA	fips
0	AK	Anchorage Borough	2	20	95010	156	ANCHORAGE (AK) ...	Anchorage AK	02020
1	AK	Kenai Peninsula Borough	2	122	17600	156	ANCHORAGE (AK) ...	Anchorage AK	02122
2	AK	Matanuska-Susitna Borough	2	170	20130	156	ANCHORAGE (AK) ...	Anchorage AK	02170
3	AK	Fairbanks North Star Borough	2	90	31860	203	FAIRBANKS (AK) ...	Fairbanks AK	02090
4	AK	Juneau Borough	2	110	25270	206	JUNEAU (AK) ...	Juneau AK	02110
...	...	...	...	...	...	...	...	...	...
3115	WY	Weston	56	45	2460	175	RAPID CITY (SD) ...	Rapid City SD	56045
3116	WY	Lincoln	56	23	5240	36	SALT LAKE CITY (UT) ...	Salt Lake City UT	56023
3117	WY	Sublette	56	35	2260	36	SALT LAKE CITY (UT) ...	Salt Lake City UT	56035
3118	WY	Sweetwater	56	37	13530	36	SALT LAKE CITY (UT) ...	Salt Lake City UT	56037
3119	WY	Uinta	56	41	6800	36	SALT LAKE CITY (UT) ...	Salt Lake City UT	56041

3120 rows × 9 columns

Now for each county we add the search volume in its DMA:

search_eclipse_2024 = []
for iii in range(len(google_dma)):
    dma_here = google_dma['GOOGLE_DMA'][iii]
    value_here = df_eclipse_2024['volume'][ df_eclipse_2024['Zone']==dma_here  ]
    try:
        search_eclipse_2024.append( int(value_here) )
    except:
        search_eclipse_2024.append( None )    
google_dma['eclipse_2024'] = search_eclipse_2024

We need to download the shape of each county.

Here we load a GeoJSON file containing the geometry information for US counties, given as a list of polygons, where feature.id is a FIPS code.

from urllib.request import urlopen
import json
with urlopen('https://raw.githubusercontent.com/plotly/datasets/master/geojson-counties-fips.json') as response:
    counties = json.load(response)

And use this in a map. choropleth_mapbox displays it on top of a Mapbox map.

import plotly
plotly.offline.init_notebook_mode()

The last two lines are not necessary, but adding them allows you to save the notebook as a standalone HTML file via File --> Download Notebook as... --> Export Notebook to HTML.

import plotly.express as px

fig = px.choropleth_mapbox(google_dma, geojson=counties, locations='fips', color='eclipse_2024',
                           color_continuous_scale="Greys",
                           range_color=(20, 80),
                           mapbox_style="carto-positron", # use "white-bg" for blank
                           zoom=3, center = {"lat": 37.0902, "lon": -95.7129},
                           opacity=1.,
                          )
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()

It can also be done on top of a world map, in any projection (see documentation https://plotly.github.io/plotly.py-docs/generated/plotly.express.choropleth.html). Using scope='usa' produces this view:

import plotly.express as px

fig = px.choropleth(google_dma, geojson=counties, locations='fips', color='eclipse_2024',
                           color_continuous_scale="Greys",
                           range_color=(20, 80),
                           scope='usa', 
                          )
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
fig.show()

More examples at https://plotly.com/python/choropleth-maps/