Silicon Valley (a term popularized by Don Hoeffler) is a region of California known for the Semiconductor Industry. Industrial semiconducor production started in the 1960’s in the Santa Clara Valley, by companies in San Jose, Santa Clara, Cupertino, Sunnyvale, Mountain View and Palo Alto. Stanford University, Hewlett-Packard, and Fairchild Semiconductor played huge roles in the start and growth of this region.
The Santa Clara Valley, mosly rural and famous for growing fruit and garlic, had the stable foundations required for vibration-sensitive wafer lithography and wafer test equipment. Densly populated urban areas like San Francisco were not suitable for this work. San Francisco was not considered part of Silicon Valley in the early days.
What changed was the booming software industry, which does not have the same physical constraints required for manufacturing of semiconductors. Software heated up in the late 1990s in many places unsuitable for silicon work, such as downtown San Francisco or lower Manhattan in New York City. Silicon Valley was initially defined by companies who develop silicon based semiconductors. In recent years the definition has expanded to include software and many other technologies.
This brings us to 2025, where Silicon Valley as a region has expanded to include the entire San Francisco/San Jose region. Some people even apply this term to technology companies no matter where they are located, however many places have their own acroynm containing the word silicon, such as Silicon Alley (NYC), Silcon Prarie (Austin), etc.
I enjoyed using map_libre again for a Figure Friday project. I have included the 58 out of 172 companies (34%) in this week’s data set based that are based in the San Francisco bay area, mostly to look at who these companies are, where are they located, and what do they do. With Map Libre, it is easy to see the 3 primary locations where these companies are located: In Santa Clara County (the original Silicon Valley), in downtown San Francisco, and in San Mateo County.
The code includes a function to wrap hover text across multiple lines. This is very useful for long descriptive strings. I used Chat GPT to make a table with longitude and latitude of these companies, and stored them in this file: “df_norcal_coords.csv”.
The code for this visualization is posted here on plotly community, however to run it you need the file with longitude and latitude values. You can get everything from:
• Git Repo: GitHub - Mike-Purtell/Plotly_FF_2024: Plotly FIgure Friday work in 2024
• Folder: Week_52_Stocks.
Here is a map_libre view of the Bay Area, showing the locations of Bay Area companies in the dataset. The marker sizes and colors are tied to Market Cap.
Here is a view of software companies clustered in downtown San Francisco.
Here is the code:
'''
Plotly Figure Friday - 2024 week 52 - stocks
58 out of 172 SaaS (Software as a Service) companies in this dataset (34%) are
based in the San Francisco Bay Area, a region that includes San Jose, Silicon
Valley, Palo Alto and Berkeley. This script uses map Libre to map the locations
of the SF Bay Area companies, with useful hover and marker sized by Market Cap.
'''
import plotly.express as px
import polars as pl
currency_to_floats = ['Stock Price']
currency_to_mils = [
'Market Cap', 'Last Quarter Revenue','Annualized Revenue',
'Last Quarter EBITDA','Annualized EBITDA','Last Quarter Net Income',
'Annualized Net Income','Cash and Short Term Investments'
]
def wrap_hover(text, CHARS_PER_LINE=45):
'''
break long hover text into multiple lines, split with html line feeds.
1st whitespace after chars_per_line value is exceeeded is replaced with <br>
'''
result = []
# Counter to track line_Length
line_length = 0
# Iterate over each character in the text
for char in text:
line_length += 1
if char.isspace():
if line_length > CHARS_PER_LINE:
result.append('<br>')
line_length = 0
else:
result.append(char)
else:
result.append(char)
return ''.join(result)
df_norcal = (
pl.scan_csv('SaaS-businesses-NYSE-NASDAQ.csv') # scan_csv --> lazy frame
.filter(pl.col('Headquarters').str.contains('California'))
.with_columns(
Headquarters =
pl.col('Headquarters')
.str.split(by =',')
.list.slice(0,1)
.list.first()
)
.filter( # exclude 4 cities on the list from southern california
~pl.col('Headquarters')
.is_in(['Glendale', 'San Diego', 'Santa Barbara','Ventura'])
)
# remove Orcacle, and Snowflake - they have moved out of the Bay Area
.filter(~pl.col('Company').str.contains('Oracle'))
.filter(~pl.col('Company').str.contains('Snowflake'))
# convert dollars as strings to millions of dollars as floats
.with_columns(pl.col(currency_to_mils).str.replace_all(',', ''))
.with_columns(pl.col(currency_to_mils).str.replace('$', '', literal=True))
.with_columns(pl.col(currency_to_mils).cast(pl.Float64))
.with_columns(pl.col(currency_to_mils)/1000000)
.with_columns(pl.col(currency_to_mils).round(0))
# convert dollars as strings to floats
.with_columns(pl.col(currency_to_floats).str.replace_all(',', ''))
.with_columns(pl.col(currency_to_floats).str.replace('$', '', literal=True))
.with_columns(pl.col(currency_to_floats).cast(pl.Float64))
.with_columns(pl.col(currency_to_floats).round(2))
# create a categorical column based on Market Cap
.with_columns(MKT_CAP_CAT = pl.lit('')) # intialize
.with_columns(
MKT_CAP_CAT =
pl.when(pl.col('Market Cap')>100e3).then(pl.lit('> $100B'))
.when(pl.col('Market Cap')>10e3).then(pl.lit('> $10B'))
.when(pl.col('Market Cap')>1e3).then(pl.lit('> $1B'))
.when(pl.col('Market Cap')>100).then(pl.lit('> $100M'))
.when(pl.col('Market Cap')>10).then(pl.lit('> $10M'))
.when(pl.col('Market Cap')>1).then(pl.lit('> $1M'))
)
.with_columns(
MARKER_SIZE =
pl.when(pl.col('Market Cap')>100e3).then(pl.lit(6))
.when(pl.col('Market Cap')>10e3).then(pl.lit(5))
.when(pl.col('Market Cap')>1e3).then(pl.lit(4))
.when(pl.col('Market Cap')>100).then(pl.lit(3))
.when(pl.col('Market Cap')>10).then(pl.lit(2))
.when(pl.col('Market Cap')>1).then(pl.lit(1))
)
.with_columns(
PROD_DESC_WRAP =
pl.col('Product Description')
.map_elements(wrap_hover, return_dtype=pl.String)
)
.drop('Company Website',
'Company Investor Relations Page',
'Lead Investor(s) Pre-IPO',
'S-1 Filing',
'September 2024 Website Traffic (Estimate)',
'YoY Change in Website Traffic%',
'2023 10-K Filing',
'Product Description'
)
.collect() # collect uses lazy frame to make polars dataframe
)
for c in currency_to_mils:
df_norcal = df_norcal.rename({c: c + ' [M$]'})
#-------------------------------------------------------------------------------
# Load file with GPS coordinates -- this data is from ChatGPT
#-------------------------------------------------------------------------------
df_coords =(
pl.scan_csv('df_norcal_coords.csv')
# remove degree notation from Long, Lat strings before convert to float
.with_columns(pl.col(['Long', 'Lat']).str.replace('°N', ''))
.with_columns(pl.col(['Long', 'Lat']).str.replace('°W', ''))
.with_columns(pl.col(['Long', 'Lat']).cast(pl.Float64))
# For Longitude, have to multiply by -1 to get degrees west
.with_columns(pl.col(['Long'])*-1.0)
.collect()
)
#-------------------------------------------------------------------------------
# Join GPS coordinates with main dataset
#-------------------------------------------------------------------------------
df = (
df_norcal
.join(
df_coords,
how = 'left',
on='Company'
)
)
df_cols = df.columns
left_cols = [c for c in df_cols if c not in ['PROD_DESC_WRAP', 'Founder(s)']]
right_cols = ['PROD_DESC_WRAP', 'Founder(s)']
df = df.select(pl.col(left_cols + right_cols))
#-------------------------------------------------------------------------------
# Make a GPS Scatter Plot of Bay Area Companies
#-------------------------------------------------------------------------------
#define marker colors based on Market Cap
dict_marker_map = {
'> $1M': px.colors.qualitative.Set1[0],
'> $10M': px.colors.qualitative.Set1[5],
'> $100M': px.colors.qualitative.Set1[4],
'> $1B': px.colors.qualitative.Set1[3],
'> $10B':px.colors.qualitative.Set1[1],
'> $100B':px.colors.qualitative.Set1[2],
}
fig = px.scatter_map(
df.sort('Market Cap [M$]', descending=True),
lat='Lat',
lon='Long',
height=1000, width=800,
size='MARKER_SIZE', # 'Market Cap [M$]',
color='MKT_CAP_CAT',
color_discrete_map= dict_marker_map,
zoom=9,
map_style='carto-voyager', #open-street-map', # 'streets',
custom_data=[
'Company', # customdata[0]
'Year Founded', # customdata[1]
'IPO Year', # customdata[2]
'Headquarters', # customdata[3]
'Market Cap [M$]', # customdata[4]
'PROD_DESC_WRAP', # customdata[5]
],
title='SaaS Companies in San Francisco Bay Area/Silicon Valley'
)
#------------------------------------------------------------------------------#
# Apply hovertemplate #
#------------------------------------------------------------------------------#
fig.update_traces(
hovertemplate =
'<b>%{customdata[0]}' +
' (Founded: %{customdata[1]}, ' +
'IPO: %{customdata[2]})</b><br>' +
'%{customdata[3]}<br>' +
'Market Cap : $%{customdata[4]:,}M<br>' +
'Products : %{customdata[5]}<br>' +
'<extra></extra>'
)
fig.update_layout(
hoverlabel=dict(
bgcolor="white",
font_size=16,
font_family='arial', # 'sans-serif mono', # 'courier new',
),
legend_title='Market Cap'
)
fig.show()
If you prefer to collaborate with others on Discord, join the Plotly Discord channel.
Welcome to the Plotly community 
