import pandas as pdfrom matplotlib import pyplot as pltimport seaborn as snsimport astimport plotly.express as pximport geopandas as gpd

le = pd.read_csv("../input/buildings-amenities-all-over-india/leisure.csv")

le.head()

# remove all tags column and rename columnsle = le.drop("All_tags", axis=1)le.columns = ["id", "name", "leisure", "lo-la"]

le.head()

# set the ID as indexle.index = le["id"]le = le.drop("id", axis = 1)

le.head()

# check NA valuesle.isna().sum()

name       27143leisure        0lo-la      37876dtype: int64

le = le[le['lo-la'].notna()]le

le['name'].fillna("missing", inplace=True)

le

le["leisure"].value_counts()

park                        1798playground                   773fitness_centre               580resort                       544pitch                        497sports_centre                495fishing                      238garden                       196stadium                      159swimming_pool                155dance                         58fitness_station               56nature_reserve                52water_park                    24marina                        20slipway                       16beach_resort                  14common                        13amusement_arcade              12track                         10yes                           10outdoor_seating                9golf_course                    8recreation_ground              6club                           6bandstand                      5bowling_alley                  5hackerspace                    5bird_hide                      4adult_gaming_centre            4sauna                          4picnic_table                   3swimming_area                  3firepit                        3horse_riding                   3cultural_centre                2gym                            2hot_spring                     2indoor_play                    2wildlife_hide                  2spa                            2Park in residential area       1aquarium                       1leisure                        1ground                         1Meeting_point                  1sports_hall                    1summer_camp                    1social_club                    1yoga                           1schoolyard                     1NITTE FOOTBALL STADIUM         1quary                          1yoga_centre                    1Name: leisure, dtype: int64

plt.rcParams['font.size'] = 10.0plt.rcParams['figure.figsize'] = 20, 10ax = sns.countplot(le['leisure'], palette="Blues_r", order=le.leisure.value_counts()[:20].index)ax.set_title("Most Avaiable Leisure Places in India")# rotate the names so they fitax.set_xticklabels(ax.get_xticklabels(), rotation=40, ha="right")plt.tight_layout()plt.show()

# split coordinatescords = list(le["lo-la"])long = []lat = []for cord in cords:    set_r = ast.literal_eval(cord)    long.append(set_r[0])    lat.append(set_r[1])le["long"] = longle["lat"] = lat

le.head()

# drop the old coordinates columnle = le.drop("lo-la", axis=1)

le.head()

# basic scatter plot of placesplt.scatter(x=le["long"], y=le["lat"])plt.show()

# create and view geopandas dataframegdf = gpd.GeoDataFrame(    le, geometry=gpd.points_from_xy(le.long, le.lat))gdf

# set mapbox acces token (required for drawing an interactive map)px.set_mapbox_access_token("pk.eyJ1IjoiYmxhY2tzdWFuMTkiLCJhIjoiY2twcDdtaGc4MDZ6djJvczR0Ym9sa3pqNCJ9.gNL1mxeSmDi6hfgwxz2qRA")

# generate and show points in map (its intractive!)fig = px.scatter_geo(gdf,                    lat=gdf.geometry.y,                    lon=gdf.geometry.x,                    hover_data=["name", "leisure"],                    locationmode="country names"                    )fig.update_geos(fitbounds="locations") # zoom in to only indiafig.show()# check the project source code for the interactive version of the map