#HIDDEN
import os
try:
import folium
except ImportError as e:
print("occurs:",e)
!pip install folium
import folium
from folium import plugins
import pandas as pd
import numpy as np
import datetime
from folium.plugins import HeatMap
Esplora la visualizzazione¶
Benvenuto nel nostro coropleto interattivo basato sui dati raccolti dalle risposte degli utenti agli eventi Meetup. Premi il tasto Show Widgets per avviare la visualizzazione. Non importa se premerai quello superiore o quello inferiore, il risultato sarà lo stesso.
NB¶
Il caricamento della mappa potrebbe richiedere anche alcuni minuti, dipendentemente dal traffico.
Non scoraggiarti, cogli l'occasione per leggere le istruzioni.
Cosa posso vedere?¶
Potrai esplorare la mappa europea osservando la rappresentazione della numerosità degli eventi che hanno trattato un determinato topic. I topic sono gli argomenti trattati durante gli eventi e possono essere molto vari: da Business fino a Tech, da LGBT ad Education.
Sicuri che la vostra mappa non rappresenti semplicemente una mappa di popolazione ?¶
Sicurissimi. Ci siamo premurati di scalare i dati in nostro possesso rispetto alla popolazione residente in macroaree così che le gradazioni di colore assunte dal coropleto non siano direttamente proporzionali alla popolazione in quella zona ma rispecchino effettivamente la proporzione di trattamento di quel topic.
Come posso accedere ai contenuti ?¶
Esplorare i contenuti è molto semplice, ti basta infatti premere sull'apposito widget:
Quando avrai interagito con il menù a tendina ti si presenterà la scelta del topic da selezionare:
NB La scala di colori utilizzata nel coropleto cambia i propri limiti massimo e minimo sulla base del topic selezionato, siccome il numero di eventi che trattano ciascun topic è diverso. Per questa ragione fai attenzione alla legenda posizionata in alto a destra nel coropleto ogni volta che cambi topic. Non è possibile confrontare direttamente i risultati tra i diversi topic, essendo l'obiettivo quello di fare un confronto tra i diversi paesi relativamente a uno stesso topic.
A questo punto non ti resta altro da fare che selezionare il topic di interesse ed esplorare liberamente.
#HIDDEN
process=False
#process=True
if process:
data=pd.read_csv("export_european_capitals.csv")
data.head()
data.sort_values("ID_event").head()
print(len(data))
for line in data.itertuples():
if data.iloc[line.Index]['dist']==0:
data['dist'].at[line.Index]=7 #arbitrary for milan
data.head()
data.sort_values(["ID_event","dist"], inplace=True)
data
ddrop=data.drop_duplicates(subset="ID_event" , keep="last")
ddrop=ddrop.drop("ID_member", axis=1)
#ddrop.sort_values("ID_event")
ddrop=ddrop.reset_index()
ddrop.head()
d=data.groupby("ID_event")['dist'].quantile(0.75).reset_index()
print(d.head())
d.rename(columns={'dist':'maxdist'}, inplace=True)
d2=data.groupby("ID_event")['dist'].mean().reset_index()
d2.rename(columns={'dist':'meandist'}, inplace=True)
ddrop=ddrop.merge(d, on="ID_event")
ddrop=ddrop.merge(d2, on="ID_event")
d3=data.groupby("ID_event")['dist'].median().reset_index()
d3.rename(columns={'dist':'mediandist'}, inplace=True)
ddrop=ddrop.merge(d3, on="ID_event")
print(ddrop.head())
d1=data.groupby("ID_event")['ID_member'].count().reset_index()
print(d1.head())
d1.rename(columns={'ID_member':'count_part'}, inplace=True)
ddrop=ddrop.merge(d1, on="ID_event")
del d1
del d2
del d
del d3
del data
ddrop.head()
event_df=pd.read_csv("../../csv/struttura/event.csv")
event_df['date']=str
for line in event_df.itertuples():
try:
event_df['date'].at[line.Index]=str(datetime.datetime.fromtimestamp(int(float(line.time)/1000)).strftime('%c'))
except:
event_df['date'].at[line.Index]="Non disponibile"
#event_df.head()
ddrop=ddrop.merge(event_df, left_on="ID_event", right_on="_key")
#ddrop.head()
del event_df
topic_df=pd.read_csv("event_topic_keywords.csv")
ddrop=ddrop.merge(topic_df,left_on="ID_event", right_on="ID").reset_index().drop("level_0",1)
del topic_df
ddrop.to_csv("post_proc.csv")
else:
ddrop=pd.read_csv("data/post_proc.csv")
#ddrop1=pd.read_csv("post_proc.csv")
#ddrop.head()
#topic_df.head()
#HIDDEN
for line in ddrop.itertuples():
if line.Country=="it":
#print(line.g_city)
if line.g_city=="Milano":
ddrop['Country'].at[line.Index]="it_Milano"
if line.g_city=="Roma":
ddrop['Country'].at[line.Index]="it_Roma"
#ddrop[ddrop['Country']=="it_Milano"]
#HIDDEN
country_list=list(ddrop.Country.unique())
#print(len(country_list))
d=ddrop.groupby("Country")['ID_event'].count()
rm1_list=[]
for elem in country_list:
if d[elem] < 50:
rm1_list.append(elem)
manual_rm_list=["ca", "ge", "ru", "cy", "us", "bg", "tr", "by", "am", "rs", "sk","si", "li", "mk", "md", "ba", "gi", "ad", "hu", "hr", "ro", "no", "fi"]#out of UE borders, or just too few events in there
rm_list=list(set(rm1_list+manual_rm_list))
#print(tot_list)
for elem in rm_list:
country_list.remove(elem)
#HIDDEN
pop={
'fr':2141000,
'es':3174000,
'gb':8136000,
'nl':821752,
'se':965232,
'de':3575000,
'cz':1281000,
'pl':1765000,
'ie':544107,
'be':174383,
'lu':114303,
'no':634293,
'hu':1756000,
'dk':602481,
'ch':133115,
'it_Milano':1352000,
'at':1868000,
'fi':631695,
'pt':504718,
'ro':1836000,
'gr':3074160,
'it_Roma':2865000,
'ua':2884000,
'hr':792875
}
#HIDDEN
ddrop=ddrop.loc[ddrop['Country'].isin(country_list)]
#HIDDEN
count_df=ddrop.groupby('Country', as_index=False).count()
count_df=count_df[['Country', 'index']].reset_index()
count_df['norm_count']=0
#HIDDEN
for line in count_df.itertuples():
norm_count=(line.index/pop[line.Country])*100000
#norm_count
count_df.norm_count.at[line.Index]=norm_count
#count_df
#HIDDEN
capitals={
'fr':"Paris",
'es':"Madrid",
'gb':"London",
'nl':"Amsterdam",
'se':"Stockholm",
'de':"Berlin",
'cz':"Praga",
'pl':"Varsaw",
'ie':"Dublin",
'be':"Bruxelles",
'lu':"Luxemburg",
'no':"Oslo",
'hu':"Budapest",
'dk':"Copenaghen",
'ch':"Berna",
'it':"Rome",
'at':"Vienna",
'fi':"Helsinki",
'pt':"Porto",
'ro':"Bucarest",
'gr':"Athens",
'ua':"Kiev",
'hr':"Zagabria",
}
#HIDDEN
from ipywidgets import interact
go=True
if go:
topic_list=[]
for line in ddrop.itertuples():
#keywords=line.Keywords.split(",")
#print(keywords)
for topic in line.Keywords.split(","):
Topic=[]
topic=topic.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
Topic.append(topic)
for topic in Topic:
if topic not in topic_list:
topic_list.append(topic)
count_dict={}
#print(topic_list)
topic_list_rm=["","Pets", "Food&Drink", "Scifi", "Beliefs", "Crafts","BookClubs", "Support", "Movements", "Games", "Films", "Lifestyle",'Moms&Dads', 'Auto', 'Paranormal', "Community", "Fashion"]
for elem in topic_list_rm:
try:
topic_list.remove(elem)
except:
print(elem)
#print(topic_list)
#HIDDEN
import json
process=False
if process:
for country in count_df.Country:
#print("#####\n",country, "\n#####\n")
punt_count_dict={}
for topic in topic_list:
#print(topic)
count_topic=0
for line in ddrop.itertuples():
if line.Country==country:
keywords=line.Keywords
#print(keywords)
for word in keywords.split():
topic_curr=word.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
#Topic.append(topics)
if topic_curr==topic:
count_topic+=1
#print(count_topic)
#print(topic,count_topic)
punt_count_dict[topic]=count_topic
#print(punt_count_dict)
count_dict[country]=punt_count_dict
#print (topic)
#print (type(line.v_lat))
#heat_data.append([float(line.v_lat),float(line.v_lon)])
with open('count_topic_dict.json', 'w') as f:
json.dump(count_dict, f)
else:
with open('data/count_topic_dict.json') as f:
count_dict = json.load(f)
#count_dict
#HIDDEN
#normalization over topics
process=False
if process:
norm_count_dict={}
for country in count_dict:
normalization=pop[country]
#print(normalization)
norm_topic_dict={}
for topic in count_dict[country]:
#print(count_dict[country][topic]) #accessing single values
norm_topic_dict[topic]=int((count_dict[country][topic]/normalization)*200000)
norm_count_dict[country]=norm_topic_dict
with open('norm_count_topic_dict.json', 'w') as f:
json.dump(norm_count_dict, f)
else:
with open('data/norm_count_topic_dict.json') as f:
norm_count_dict = json.load(f)
#norm_count_dict
#HIDDEN
go=False
if go:
@interact(topic=topic_list)
def show_canned_examples(topic):
lat_df=ddrop[['v_lat', 'v_lon']]
print(len(ddrop))
str(lat_df.v_lat[0]) == "nan"
for line in lat_df.itertuples():
if str(line.v_lat) == "nan":
lat_df['v_lat'].at[line.Index]= ddrop['g_lat'].at[line.Index]
if str(line.v_lon) == "nan":
lat_df['v_lon'].at[line.Index]= ddrop['g_lon'].at[line.Index]
len(lat_df)
lat_df.isna().any()
lat_df=lat_df.astype(float)
from folium.plugins import HeatMap
map_hooray = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=5)
# List comprehension to make out list of lists
#heat_data = [[row['v_lat'],row['v_lon']] for index, row in lat_df.iterrows()]
heat_data=[]
for line in lat_df.itertuples():
if ddrop.Country.at[line.Index] in country_list:
keywords=ddrop.Keywords.at[line.Index]
#print(keywords)
for topics in keywords.split():
topic_curr=topics.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
#Topic.append(topics)
if topic_curr==topic:
#print (topic)
#print (type(line.v_lat))
heat_data.append([float(line.v_lat),float(line.v_lon)])
# Plot it on the map
HeatMap(heat_data).add_to(map_hooray)
# Display the map
map_hooray.save("/home/dario/NeoMeetup/Visualizations/map/heat_try.html")
display(map_hooray)
#HIDDEN
from ipywidgets import interact
go=False
if go:
topic_list=[]
for line in ddrop.itertuples():
#keywords=line.Keywords.split(",")
#print(keywords)
for topic in line.Keywords.split(","):
Topic=[]
topic=topic.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
Topic.append(topic)
for topic in Topic:
if topic not in topic_list:
topic_list.append(topic)
#print(topic_list)
@interact(topic=topic_list)
def show_canned_examples(topic):
lat_df=ddrop[['v_lat', 'v_lon']]
print(len(ddrop))
str(lat_df.v_lat[0]) == "nan"
for line in lat_df.itertuples():
if str(line.v_lat) == "nan":
lat_df['v_lat'].at[line.Index]= ddrop['g_lat'].at[line.Index]
if str(line.v_lon) == "nan":
lat_df['v_lon'].at[line.Index]= ddrop['g_lon'].at[line.Index]
len(lat_df)
lat_df.isna().any()
lat_df=lat_df.astype(float)
from folium.plugins import HeatMap
map_hooray = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=4)
# List comprehension to make out list of lists
#heat_data = [[row['v_lat'],row['v_lon']] for index, row in lat_df.iterrows()]
heat_data=[]
for row in count_df.itertuples():
norm_count=0
for line in lat_df.itertuples():
if ddrop.Country.at[line.Index] == row.Country:
norm_count+=1
keywords=ddrop.Keywords.at[line.Index]
#print(keywords)
for topics in keywords.split():
topic=topics.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
#Topic.append(topics)
if topic in topic_list:
#print (topic)
#print (type(line.v_lat))
heat_data.append([float(line.v_lat),float(line.v_lon)])
if norm_count==row.norm_count:
print("for country {} inserted {} events".format(row.Country, norm_count))
break
# Plot it on the map
HeatMap(heat_data).add_to(map_hooray)
plugins.Fullscreen(
position='bottomright',
title='Expand me',
title_cancel='Exit me',
force_separate_button=True
).add_to(map_hooray)
# Display the map
map_hooray.save("/home/dario/NeoMeetup/Visualizations/map/heat_try.html")
#heat_dict[]
display(map_hooray)
#HIDDEN
lat_df=ddrop[['v_lat', 'v_lon']]
#print(len(ddrop))
str(lat_df.v_lat[0]) == "nan"
for line in lat_df.itertuples():
if str(line.v_lat) == "nan":
lat_df['v_lat'].at[line.Index]= ddrop['g_lat'].at[line.Index]
if str(line.v_lon) == "nan":
lat_df['v_lon'].at[line.Index]= ddrop['g_lon'].at[line.Index]
len(lat_df)
lat_df.isna().any()
lat_df=lat_df.astype(float)
#HIDDEN
process=False
if process:
heat_df = pd.DataFrame(columns=["topic", "v_lat", "v_lon", "Country"])
topic_map_dict={}
for topic in topic_list:
print("#######\ntopic is ", topic, "\n######")
map_hooray = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=5)
# List comprehension to make out list of lists
#heat_data = [[row['v_lat'],row['v_lon']] for index, row in lat_df.iterrows()]
heat_data=[]
for country in count_df.Country:
norm_count=0
for line in lat_df.itertuples():
if ddrop.Country.at[line.Index] == country:
keywords=ddrop.Keywords.at[line.Index]
#print(keywords)
for topics in keywords.split():
curr_topic=topics.replace("[","").replace("]","").replace("'","").replace(",","").replace(" ","")
#Topic.append(topics)
if curr_topic == topic:
#print ("appending")
#print (type(line.v_lat))
#heat_data.append([float(line.v_lat),float(line.v_lon)])
heat_df=heat_df.append({"topic":topic,
"v_lat":line.v_lat,
"v_lon":line.v_lon,
"Country":country}, ignore_index=True)
norm_count+=1
#print(norm_count)
#print
if norm_count==norm_count_dict[country][topic]:
print("for country {} inserted {} events".format(country, norm_count))
break
if norm_count==norm_count_dict[country][topic]:
break
#print("for country {} inserted {} events".format(country, norm_count))
heat_df.to_csv("heat_df.csv", index=False)
heat_df.head()
#HIDDEN
go=False
if go:
from folium.plugins import HeatMap
topic_map_dict={}
for topic in topic_list:
map_hooray = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=5)
heat_data=[]
for country in count_df.Country:
norm_count=0
for line in heat_df.itertuples():
if line.Country == country:
if line.topic == topic:
heat_data.append([float(line.v_lat),float(line.v_lon)])
# Plot it on the map
HeatMap(heat_data).add_to(map_hooray)
plugins.Fullscreen(
position='bottomright',
title='Expand me',
title_cancel='Exit me',
force_separate_button=True
).add_to(map_hooray)
# Display the map
topic_map_dict[topic]=map_hooray
@interact(topic=topic_list)
def show_canned_examples(topic):
return topic_map_dict[topic]
#HIDDEN
import warnings
warnings.filterwarnings('ignore')
try:
heat_df=pd.read_csv("heat_df.csv")
except:
heat_df=pd.read_csv("https://raw.githubusercontent.com/DBertazioli/NBinteract/master/heat_df.csv")
from folium.plugins import HeatMap
topic_map_dict={}
go=False
if go:
@interact(topic=topic_list)
def show_canned_examples(topic):
map_hooray = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=5)
heat_data=[]
for country in count_df.Country:
norm_count=0
for line in heat_df.itertuples():
if line.Country == country:
if line.topic == topic:
heat_data.append([float(line.v_lat),float(line.v_lon)])
# Plot it on the map
HeatMap(heat_data).add_to(map_hooray)
plugins.Fullscreen(
position='bottomright',
title='Expand me',
title_cancel='Exit me',
force_separate_button=True
).add_to(map_hooray)
# Display the map
return map_hooray
heat_df.head()
count_try=count_df
count_try=count_try.drop(11).drop("level_0", axis=1).reset_index().drop("level_0", axis=1)
count_try["cloro_id"]="NaN"
count_try
id_list=["AT","BE","CH","CZ","DE","DK","ES","FR","GB","GR", "IE", "IT", "LU","NL","PL","PT","SE","UA"]
for i in range(len(count_try)):
#print(id_list[i])
count_try["cloro_id"][i]=id_list[i]
#count_try["cloro_id"][1]="BE"
#count_try["cloro_id"][0]="AT"
count_try
count_dict
id_list=["AT","BE","CH","CZ","DE","DK","ES","FI","FR","GB","GR","HR","HU", "IE", "IT", "LU","NL","NO","PL","PT","RO","SE","UA"]
topic_df_dict={}
for topic in topic_list:
df = pd.DataFrame(columns=['Country','Count'])#, index=['x','y','z'])
for it,key in enumerate(norm_count_dict.keys()):
#print(key)
if norm_count_dict[key][topic]*10==1530:
count=1500
elif norm_count_dict[key][topic]*10-10 <= 0:
count=0
else:
count=norm_count_dict[key][topic]*10-10
df = df.append({
"Country": key,
"Count": count
}, ignore_index=True)
df=df.drop(14).reset_index()
df["cloro_id"]="NaN"
for i in range(len(df)):
#print(id_list[i])
df["cloro_id"][i]=id_list[i]
topic_df_dict[topic]=df
#print(topic_df_dict[topic])
#print(topic)
#print(df)
#topic_df_dict
#for key in topic_df_dict.keys():
# print(key)
#topic_df_dict["Business"]
#HIDDEN
topic_list.sort()
@interact(topic=topic_list)
def show_canned_examples(topic):
map1 = folium.Map(location=[ddrop['v_lat'].mean(),ddrop['v_lon'].mean()], zoom_start=4, tiles="cartodbpositron")
folium.Choropleth(
#geo_data="/home/dario/Downloads/custom.geo(1).json",
geo_data="https://raw.githubusercontent.com/DBertazioli/Interact/master/custom.json",
data=topic_df_dict[topic],
columns=['cloro_id', 'Count'],
key_on='feature.properties.iso_a2',
fill_color='BuPu',
legend_name="number of events").add_to(map1)
# Plot it on the map
plugins.Fullscreen(
position='bottomright',
title='Expand me',
title_cancel='Exit me',
force_separate_button=True
).add_to(map1)
print("Per le regioni in nero non si dispone di dati riguardo ai Meetups")
# Display the map
return map1
The map is built using folium API to leaflet.
View the full code here.