Analise_Exploratoria_Dados/aed_python.py at main · malucor/Analise_Exploratoria_Dados · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
import json
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
import pandas as pd
import geopandas
import geopy
from geopy.geocoders import Nominatim
from geopy.extra.rate_limiter import RateLimiter

# ---------------
# Coleta de Dados

wget -q "https://raw.githubusercontent.com/andre-marcos-perez/ebac-course-utils/main/dataset/deliveries.json" -O deliveries.json
wget -q "https://raw.githubusercontent.com/andre-marcos-perez/ebac-course-utils/main/dataset/deliveries-geodata.csv" -O deliveries-geodata.csv
wget -q "https://geoftp.ibge.gov.br/cartas_e_mapas/bases_cartograficas_continuas/bc100/go_df/versao2016/shapefile/bc100_go_df_shp.zip" -O distrito-federal.zip
unzip -q distrito-federal.zip -d ./maps
cp ./maps/LIM_Unidade_Federacao_A.shp ./distrito-federal.shp
cp ./maps/LIM_Unidade_Federacao_A.shx ./distrito-federal.shx

# ----------------------
# Wrangling da Estrutura

with open('deliveries.json', mode='r', encoding='utf8') as file:
  data = json.load(file)

deliveries_df = pd.DataFrame (data)
deliveries_df.head()

hub_origin_df = pd.json_normalize(deliveries_df["origin"])
hub_origin_df.head()

deliveries_df = pd.merge(left=deliveries_df, right=hub_origin_df, how='inner', left_index=True, right_index=True)
deliveries_df.head()

deliveries_df = deliveries_df.drop("origin", axis=1)
deliveries_df = deliveries_df[["name", "region", "lng", "lat", "vehicle_capacity", "deliveries"]]
deliveries_df.head()

deliveries_df.rename(columns={"lng": "hub_lng", "lat": "hub_lat"}, inplace=True)
deliveries_df.head()

deliveries_df.rename(columns={"lng": "hub_lng", "lat": "hub_lat"}, inplace=True)
deliveries_df.head()

deliveries_normalized_df = pd.concat([
    pd.DataFrame(deliveries_exploded_df["deliveries"].apply(lambda record: record["size"])).rename(columns={"deliveries": "delivery_size"}),
    pd.DataFrame(deliveries_exploded_df["deliveries"].apply(lambda record: record["point"]["lng"])).rename(columns={"deliveries": "delivery_lng"}),
    pd.DataFrame(deliveries_exploded_df["deliveries"].apply(lambda record: record["point"]["lat"])).rename(columns={"deliveries": "delivery_lat"}),
], axis= 1)
deliveries_normalized_df.head()

deliveries_df = deliveries_df.drop("deliveries", axis=1)
deliveries_df = pd.merge(left=deliveries_df, right=deliveries_normalized_df, how='right', left_index=True, right_index=True)
deliveries_df.reset_index(inplace=True, drop=True)
deliveries_df.head()

# --------------------
# Exploração do Schema

deliveries_df.dtypes
deliveries_df.select_dtypes("object").describe().transpose()
deliveries_df.drop(["name", "region"], axis=1).select_dtypes('int64').describe().transpose()

# ---------------
# Dados Faltantes

deliveries_df.isna().any()

# --------------
# Enriquecimento

hub_df = deliveries_df[["region", "hub_lng", "hub_lat"]]
hub_df = hub_df.drop_duplicates().sort_values(by="region").reset_index(drop=True)
hub_df.head()

geolocator = Nominatim(user_agent="ebac_geocoder")
location = geolocator.reverse("-15.657013854445248, -47.802664728268745")

print(json.dumps(location.raw, indent=2, ensure_ascii=False))

geocoder = RateLimiter(geolocator.reverse, min_delay_seconds=1)

hub_df["coordinates"] = hub_df["hub_lat"].astype(str) + ", " + hub_df["hub_lng"].astype(str)
hub_df["geodata"] = hub_df["coordinates"].apply(geocoder)
hub_df.head()

hub_geodata_df = pd.json_normalize(hub_df["geodata"].apply(lambda data: data.raw))
hub_geodata_df.head()

hub_geodata_df = hub_geodata_df[["address.town", "address.suburb", "address.city"]]
hub_geodata_df.rename(columns={"address.town": "hub_town", "address.suburb": "hub_suburb", "address.city": "hub_city"}, inplace=True)
hub_geodata_df["hub_city"] = np.where(hub_geodata_df["hub_city"].notna(), hub_geodata_df["hub_city"], hub_geodata_df["hub_town"])
hub_geodata_df["hub_suburb"] = np.where(hub_geodata_df["hub_suburb"].notna(), hub_geodata_df["hub_suburb"], hub_geodata_df["hub_city"])
hub_geodata_df = hub_geodata_df.drop("hub_town", axis=1)
hub_geodata_df.head()

hub_df = pd.merge(left=hub_df, right=hub_geodata_df, left_index=True, right_index=True)
hub_df = hub_df[["region", "hub_suburb", "hub_city"]]
hub_df.head()

deliveries_df = pd.merge(left=deliveries_df, right=hub_df, how="inner", on="region")
deliveries_df = deliveries_df[["name", "region", "hub_lng", "hub_lat", "hub_city", "hub_suburb", "vehicle_capacity", "delivery_size", "delivery_lng", "delivery_lat"]]
deliveries_df.head()

deliveries_geodata_df = pd.read_csv("deliveries-geodata.csv")
deliveries_geodata_df.head()

deliveries_df = pd.merge(left=deliveries_df, right=deliveries_geodata_df[["delivery_city", "delivery_suburb"]], how="inner", left_index=True, right_index=True)
deliveries_df.head()

# ---------------------
# Controle de Qualidade

deliveries_df.info()
deliveries_df.isna().any()

100 * (deliveries_df["delivery_city"].isna().sum() / len(deliveries_df))
100 * (deliveries_df["delivery_suburb"].isna().sum() / len(deliveries_df))

prop_df = deliveries_df[["delivery_city"]].value_counts() / len(deliveries_df)
prop_df.sort_values(ascending=False).head(10)

prop_df = deliveries_df[["delivery_suburb"]].value_counts() / len(deliveries_df)
prop_df.sort_values(ascending=False).head(10)

# ---------------------
# Visualização com Mapa

mapa = geopandas.read_file("distrito-federal.shp")
mapa = mapa.loc[[0]]
mapa.head()

hub_df = deliveries_df[["region", "hub_lng", "hub_lat"]].drop_duplicates().reset_index(drop=True)
geo_hub_df = geopandas.GeoDataFrame(hub_df, geometry=geopandas.points_from_xy(hub_df["hub_lng"], hub_df["hub_lat"]))
geo_hub_df.head()

geo_deliveries_df = geopandas.GeoDataFrame(deliveries_df, geometry=geopandas.points_from_xy(deliveries_df["delivery_lng"], deliveries_df["delivery_lat"]))
geo_deliveries_df.head()

fig, ax = plt.subplots(figsize = (50/2.54, 50/2.54))

mapa.plot(ax=ax, alpha=0.4, color="lightgrey")

geo_deliveries_df.query("region == 'df-0'").plot(ax=ax, markersize=1, color="crimson", label="df-0")
geo_deliveries_df.query("region == 'df-1'").plot(ax=ax, markersize=1, color="teal", label="df-1")
geo_deliveries_df.query("region == 'df-2'").plot(ax=ax, markersize=1, color="darkgoldenrod", label="df-2")

geo_hub_df.plot(ax=ax, markersize=30, marker="x", color="black", label="hub")

plt.title("Entregas no Distrito Federal por Região", fontdict={"fontsize": 16})
lgnd = plt.legend(prop={"size": 15})
for handle in lgnd.legendHandles:
  handle.set_sizes([50])

# ----------------------------------
# Visualização com Gráfico de Barras

data = pd.DataFrame(deliveries_df[['region', 'vehicle_capacity']].value_counts(normalize=True)).reset_index()
data.rename(columns={0: "region_percent"}, inplace=True)
data.head()

with sns.axes_style('whitegrid'):
  grafico = sns.barplot(data=data, x="region", y="region_percent", errorbar=None, palette='Blues_r')
  grafico.set(title='Proporção de Entregas por Região', xlabel='Região', ylabel='Proporção');