import altair as alt
from vega_datasets import data
import geopandas as gpd
df_quakes = gpd.read_file("lastday.json")
df_quakes = df_quakes[df_quakes["mag"] != "-"]
df_quakes["mag_num"] = df_quakes["mag"].astype(float)
df_quakes = df_quakes[df_quakes.mag_num > 0]
# Data generators for the background
sphere = alt.sphere()
graticule = alt.graticule()
# Source of land data
# source = alt.topo_feature(data.world_110m.url, 'countries')
source = gpd.read_file("AT_and_neighbors.geojson")
# Layering and configuring the components
background = alt.Chart(source).mark_geoshape(fill='lightgrey', stroke='white')
points = alt.Chart(df_quakes).mark_circle().encode(
longitude="lon:Q",
latitude="lat:Q",
#size=alt.Size("mag_num:Q", title="Magnitude"),
#size="magmag:Q",
color="red",
fillOpacity=0.5
# ).transform_calculate(
# magmag="int(10*datum.mag)"
).project(
type='conicConformal',
center=[8, -3],
# --- Graticule Generator
import altair as alt
data = alt.graticule(step=[15, 15])
alt.Chart(data).mark_geoshape(stroke='black').project(
'orthographic',
rotate=[0, -45, 0]
)
# --- Sphere Generator
import altair as alt
sphere_data = alt.sphere()
grat_data = alt.graticule(step=[15, 15])
background = alt.Chart(sphere_data).mark_geoshape(fill='aliceblue')
lines = alt.Chart(grat_data).mark_geoshape(stroke='lightgrey')
alt.layer(background, lines).project('naturalEarth1')
######## ---------
#
# Encodings
import altair as alt
from vega_datasets import data
cars = data.cars()
def write():
st.header("FIFA Players Across the World")
st.write("""Let's explore the distribution of footballers around the world.
Hover over the map to see player stats for different countries
grouped by player nationality.""")
# Load in the data
fifa_country_agg = load_data('data/clean_fifa_country_aggs.csv')
continents_zoom_params = load_continent_zoom_params()
continents = list(continents_zoom_params.keys())
# Add sliders and checkboxes for users to configure visualization
player_stats_dict, player_stats_labels, player_stats_columns = load_player_stats(
)
agg_functions = ['Mean', 'Min', 'Max']
# Show results for top N countries
num_countries = fifa_country_agg.shape[0]
st.markdown('### Show Top N Countries')
top_countries_count = st.slider('N', 0, num_countries, num_countries)
# Select continent to zoom into on map
select_continent = st.selectbox('Continent',
options=continents,
index=continents.index('All'))
st.subheader("Map: %s" % select_continent)
# Select which player stats to show
st.sidebar.markdown('### Show Player Stats')
show_player_stats = st.sidebar.multiselect(
'Show in Tooltip and Table',
options=player_stats_labels,
default=['Age', 'Overall Rating (0-100)', 'Potential Rating (0-100)'])
# Select which aggregate functions to include
st.sidebar.markdown('### Show Aggregate Functions')
agg_functions_checkbox = {}
for a in agg_functions:
agg_functions_checkbox[a] = st.sidebar.checkbox(a, value=a in ['Mean'])
# Select how to determine top N countries
st.sidebar.markdown('### Top N Countries Based On')
top_countries_attr = st.sidebar.selectbox('Player Stat',
options=player_stats_labels,
index=3)
top_countries_agg = st.sidebar.selectbox('Using', options=agg_functions)
top_countries_order = st.sidebar.selectbox(
'Order', options=['Ascending', 'Descending'], index=1)
# Draw the world map of FIFA 19 player nationalities
# source: https://altair-viz.github.io/gallery/index.html#maps
# Data to show based on user selections
show_df = fifa_country_agg.sort_values(
by='%s_%s' %
(player_stats_dict.get(top_countries_attr), top_countries_agg.lower()),
ascending=top_countries_order == 'Ascending').head(top_countries_count)
# Data generators for the background
sphere = alt.sphere()
graticule = alt.graticule()
# Source of land data
source = alt.topo_feature(data.world_110m.url, 'countries')
# Layering and configuring the components
background = alt.layer(
alt.Chart(sphere).mark_geoshape(fill='lightblue'),
alt.Chart(graticule).mark_geoshape(stroke='white', strokeWidth=0.2),
alt.Chart(source).mark_geoshape(
fill='#9eb5a8', stroke='black')).project(
type='equirectangular', # map type
scale=continents_zoom_params.get(select_continent)[0],
center=continents_zoom_params.get(select_continent)
[1:3]).properties(width=800,
height=400).configure_view(stroke=None)
hover = alt.selection(type='single',
on='mouseover',
nearest=True,
fields=['Latitude', 'Longitude'])
# Get fields to show in tooltip
tooltip_info = ['Nationality Country']
for player_stat_label in show_player_stats:
player_stat_column = player_stats_dict.get(player_stat_label)
for agg_function, function_checked in agg_functions_checkbox.items():
if function_checked:
tooltip_info.append("%s_%s" %
(player_stat_column, agg_function.lower()))
base = alt.Chart(show_df).encode(longitude='Longitude:Q',
latitude='Latitude:Q',
tooltip=tooltip_info)
points = base.mark_point().encode(
color=alt.condition(~hover, alt.value('#014600'), alt.value('red')),
size=alt.condition(~hover, alt.value(30),
alt.value(100))).add_selection(hover)
st.write(background + points)
st.subheader("Data Shown on Map")
st.write(show_df[tooltip_info])
def plot_all(dataset=f"data/{_OUTPUT_JSON_FILENAME}"):
"""
Function to plot all available maps and charts that interact.
Available charts and maps:
plot_stats_world
plot_tweets_world
plot_stats_usa
plot_tweets_usa
plot_choro_usa
Args:
dataset: JSON formatted file with tweet data.
Defaults to "data/_OUTPUT_JSON_FILENAME".
Returns:
output_map: Collection of all maps and charts.
"""
try:
# calling utility functions to create dataframes
df_world = create_dataframe_world(dataset)
df_usa = create_dataframe_us(dataset)
choro = create_dataframe_choropleth(dataset)
# using vega_datasets(data)
countries = alt.topo_feature(data.world_110m.url, feature='countries')
states = alt.topo_feature(data.us_10m.url, feature="states")
# using altair function to create a spherical background
sphere = alt.sphere()
# setting variables for interaction
# setting actual selection
brush = alt.selection_multi(encodings=["y"])
# setting the opacity of (un)selected bars
opacity_overview = alt.condition(brush, alt.value(0.9), alt.value(0.1))
# setting the opacity of (un)selected points
opacity_points = alt.condition(brush, alt.value(1), alt.value(0))
# disable the max rows restriction to be able to chart larger datasets
alt.data_transformers.disable_max_rows()
# spheric background
spheric_background = alt.Chart(sphere).mark_geoshape(fill="aliceblue")
# us-states background
background_usa = alt.Chart(states).mark_geoshape(
fill="lightgray",
stroke="white"
).properties(
width=720,
height=400
)
# world countries background
background_world = alt.Chart(countries).mark_geoshape(
fill='lightgray',
stroke='white'
).properties(
width=720,
height=400
)
# plot_tweets_world()
tweets_world = alt.Chart(df_world).mark_square(
size=1, opacity=0.8).encode(
longitude="coord_lon",
latitude="coord_lat",
color=alt.Color("sentiment:N",
scale=alt.Scale(
domain=["positive", "neutral", 'negative'],
range=["green", "orange", "red"],
)
),
tooltip=["user_name", "text", "sentiment"],
opacity=opacity_points
).add_selection(brush).properties(title="Tweets of the world")
# plot_tweets_usa()
tweets_usa = alt.Chart(df_usa).mark_square(size=5, opacity=0.5).encode(
longitude="coord_lon",
latitude="coord_lat",
color=alt.Color("sentiment:N",
scale=alt.Scale(
domain=["positive", "neutral", "negative"],
range=["green", "orange", "red"]
)
),
tooltip=["user_name", "text", "sentiment"],
opacity=opacity_points
).add_selection(brush).properties(title="Tweets of the US")
# plot_stats_world()
stats_world = alt.Chart(df_world).mark_bar().encode(
alt.Y("source_clean:N",
title="sources"
),
alt.X("count()"),
opacity=opacity_overview,
tooltip="count()"
).add_selection(
brush
).properties(
width=720,
title="Tweet sources of the world"
).interactive()
# plot_stats_usa()
stats_usa = alt.Chart(df_usa).mark_bar().encode(
alt.Y("source_clean:N",
title="sources"
),
alt.X("count()"),
opacity=opacity_overview,
tooltip="count()"
).add_selection(
brush
).properties(
width=720,
title="Tweet sources of the US"
).interactive()
# plot_choro_usa()
choro = alt.Chart(choro).mark_geoshape().encode(
color=alt.Color('sentiment',
legend=alt.Legend(title="Positivity",
tickCount=5,
tickMinStep=0.5),
scale=alt.Scale(
scheme='redyellowgreen', domain=[-1, 1])
),
tooltip=["name:N", "sentiment"]
).properties(
width=720,
height=400,
title="Average tweet sentiment per state of the US"
).add_selection(brush)
# generating the final layered chart
output_map = stats_world &\
(spheric_background + background_world + tweets_world).project(
"naturalEarth1") &\
stats_usa &\
(background_usa + tweets_usa).project("albersUsa") &\
(background_usa + choro).project("albersUsa")
return output_map
except Exception:
raise ValueError("""There are no tweets in the dataset yet!
Try running get_data() again to collect more data or
select another dataset.""")
def plot_tweets_world(dataset=f"data/{_OUTPUT_JSON_FILENAME}"):
"""
Function to plot all tweets worldwide as points using altair.
Args:
dataset: JSON formatted file with tweet data.
Defaults to "data/_OUTPUT_JSON_FILENAME".
Returns:
output_map: map that represents the world and all tweets.
Raises:
ValueError: If no tweets can be found in the dataset.
"""
try:
# calling utility function to create dataframe
df = create_dataframe_world(dataset)
# using vega_datasets(data) to get the borders of the countries
countries = alt.topo_feature(data.world_110m.url, feature='countries')
# using altair function to create a spherical background
sphere = alt.sphere()
# disable the max rows restriction to be able to chart large datasets
alt.data_transformers.disable_max_rows()
# generating chart consisting of the spherical background
spheric_background = alt.Chart(sphere).mark_geoshape(fill="aliceblue")
# generating chart consisting of the countries
background_world = alt.Chart(countries).mark_geoshape(
fill='lightgray',
stroke='white'
).properties(
width=720,
height=400
)
# generating chart consisting of the tweets
points_world = alt.Chart(df).mark_square(size=1, opacity=0.8).encode(
# setting the lat/lon
longitude="coord_lon",
latitude="coord_lat",
color=alt.Color("sentiment:N",
# matching the classes onto the sentiment
scale=alt.Scale(domain=[
"positive", "neutral", 'negative'
],
# colouring the classes
range=["green", "orange", "red"]
)
),
# generating tooltip with the user name, text, sentiment of tweet
tooltip=["user_name", "text", "sentiment"]
).properties(title="Tweets of the world")
# generating the final layered chart
output_map = (spheric_background + background_world + points_world)\
.project("naturalEarth1")
return output_map
except Exception:
raise ValueError("""There are no tweets in the dataset yet!
Try running get_data() again to collect more data or
select another dataset.""")
def generate_countries_map(data: pd.DataFrame,
date,
interactive: bool = False,
width: int = 600,
height: int = 600,
log_scale: bool = True) -> alt.Chart:
fechas = data['fecha'].apply(lambda x: x.date())
data = data[(fechas == date)]
scale = alt.Scale(type='log', scheme='teals') if log_scale else alt.Scale(
type='linear', scheme='teals')
url_country_name = 'https://raw.githubusercontent.com/alisle/world-110m-country-codes/master/world-110m-country-codes.json'
country_names = pd.read_json(url_country_name).drop('name', axis=1)
country_data = get_country_data()
country_data = country_names.join((country_data.set_index('code')),
on='code')
data = pd.merge(left=country_data,
right=data,
left_on='name',
right_on='pais').dropna()
data = data.astype({'id': int, 'casos': int})
sphere = alt.sphere()
graticule = alt.graticule()
source = alt.topo_feature(vd.data.world_110m.url, 'countries')
sphere_chart = alt.Chart(
sphere,
title='Ubicación de los casos confirmados por país').mark_geoshape(
fill='lightblue')
graticule_chart = alt.Chart(graticule).mark_geoshape(stroke='white',
strokeWidth=0.5)
countries_chart = (alt.Chart(source).mark_geoshape().encode(
color=alt.Color('casos:Q', title='Casos', scale=scale, legend=None),
tooltip=[
alt.Tooltip('name:N', title='País'),
alt.Tooltip('code:N', title='Código'),
alt.Tooltip('casos:Q', title='Casos')
]).transform_lookup('id',
from_=alt.LookupData(
data=data,
key='id',
fields=['code', 'name', 'casos'])))
single = alt.selection_single(
on='mouseover', nearest=True, fields=['pais'],
empty='all') if interactive else alt.selection_single()
circle_chart = (alt.Chart(source).mark_circle(
opacity=0.4, color='red').encode(
longitude='lon:Q',
latitude='lat:Q',
size=(alt.condition(
single,
alt.Size('casos:Q',
scale=alt.Scale(range=[50, 4000]),
legend=None), alt.value(0))),
tooltip=[
alt.Tooltip('pais:N', title='País'),
alt.Tooltip('code:N', title='Código'),
alt.Tooltip('casos:Q', title='Casos')
]).transform_lookup('id',
from_=alt.LookupData(data=data,
key='id',
fields=[
'code', 'pais',
'casos', 'lat', 'lon'
])).add_selection(single))
final_chart = ((sphere_chart + graticule_chart + countries_chart +
circle_chart).project('naturalEarth1').properties(
width=800, height=500).configure_view(stroke=None))
return final_chart
