-
Notifications
You must be signed in to change notification settings - Fork 0
/
format_data.py
218 lines (192 loc) · 8.07 KB
/
format_data.py
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
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
import pandas as pd
import numpy as np
import time
import math
import json
from tqdm import tqdm
from geopy import Nominatim
import pickle
map_geopy_dvf = {
"adresse_numero": "house_number", # split string and take first
"adresse_nom_voie": "road",
"adresse_code_voie": "house_number", # split string and take second
"code_postal": "postcode",
"code_commune": "ballec",
"nom_commune": ["town", "village"], # take town if present else take village
"code_departement": "postcode", # take first 2 digits
}
def recover_adress(df, year):
locator = Nominatim(user_agent="myGeocoder")
address_dict = {}
add = []
for latitude, longitude in df[df["address_known"] == False][
["latitude", "longitude"]
].itertuples(index=False):
if not math.isnan(latitude):
add.append((latitude, longitude))
add = set(add)
with tqdm(total=len(add)) as bar:
for latitude, longitude in add:
bar.set_description(
f"processing addresses lat={str(latitude)}, lon={longitude}"
)
try:
addresse = str(latitude) + ", " + str(longitude)
address_dict[(latitude, longitude)] = locator.reverse(addresse)
except Exception as e:
print(
"Could not find addresse for coordinates ",
str(latitude),
str(longitude),
e,
)
bar.update(1)
for k, v in address_dict.items():
try:
address_dict[k] = v.raw
except:
print("There was an issue with:", k, v)
with open(f"address_{year}.pickle", "wb") as f:
pickle.dump(address_dict, f)
return address_dict
def add_addresses_to_df(df, addresses: dict):
with tqdm(total=df[~df["address_known"]].shape[0]) as bar:
bar.set_description("Adding addresses to dataframe")
for index in df[~df["address_known"]].index:
if not math.isnan(df.loc[index, "latitude"]):
lat_long = (df.loc[index, "latitude"], df.loc[index, "longitude"])
new_addresse = addresses[lat_long]["address"]
# numéro et code d'adresse
if "house_number" in new_addresse:
numbers = new_addresse["house_number"].split(" ")
if not isinstance(df.loc[index, "adresse_numero"], str):
df.loc[index, "adresse_numero"] = numbers[0]
if (
not isinstance(df.loc[index, "adresse_code_voie"], str)
and math.isnan(df.loc[index, "adresse_code_voie"])
and len(numbers) > 1
):
df.loc[index, "adresse_code_voie"] = numbers[1]
# nom de voie
if "road" in new_addresse:
if not isinstance(df.loc[index, "adresse_nom_voie"], str):
df.loc[index, "adresse_nom_voie"] = new_addresse["road"]
# code postale et code département
if "postcode" in new_addresse:
code_postale = new_addresse["postcode"]
if len(code_postale) > 5:
code_postale = code_postale[:5]
elif len(code_postale) < 5:
while len(code_postale) < 5:
code_postale = "0" + code_postale
if math.isnan(df.loc[index, "code_postal"]):
df.loc[index, "code_postal"] = code_postale
if isinstance(df.loc[index, "code_departement"], str):
# Si c'est un domtom on prend les 3 premiers numéros, sinon 2
if code_postale[:2] == "97":
df.loc[index, "code_departement"] = code_postale[:3]
else:
df.loc[index, "code_departement"] = code_postale[:2]
# nom de la commune
if "town" in new_addresse:
if not isinstance(df.loc[index, "nom_commune"], str):
df.loc[index, "nom_commune"] == new_addresse["town"]
# parfois ya pas de commune donc le village est pris
elif "village" in new_addresse:
if not isinstance(df.loc[index, "nom_commune"], str):
df.loc[index, "nom_commune"] == new_addresse["village"]
bar.update(1)
def format_code_postal(text) -> str:
if isinstance(text, float) and not math.isnan(text):
postal_code = str(int(text))
if len(postal_code) < 5:
while len(postal_code) < 5:
postal_code = "0" + postal_code
elif len(postal_code) > 5:
postal_code = postal_code[:5]
return postal_code
elif isinstance(text, float) and math.isnan(text):
return ""
elif isinstance(text, str):
return text
def format_adresse_number(text) -> str:
if isinstance(text, float) and not math.isnan(text):
return str(int(text))
elif isinstance(text, float) and math.isnan(text):
return ""
elif isinstance(text, str):
return text
def format_valeur_fonciere(number) -> int:
if isinstance(number, float) and not math.isnan(number):
return int(number)
elif isinstance(number, float) and math.isnan(number):
return None
def open_csv_file(filename: str, delimiter: str = ","):
try:
with open(filename, "r") as f:
data = pd.read_csv(f, delimiter=delimiter)
return data
except IOError as e:
raise (e)
def clean_dataframe_DVF(df):
df["adresse_numero"] = df["adresse_numero"].apply(
lambda y: str(int(y)) if (isinstance(y, float) and not math.isnan(y)) else y
)
df["valeur_fonciere"] = df["valeur_fonciere"].apply(
lambda y: int(y) if (isinstance(y, float) and not math.isnan(y)) else y
)
# si l'adresse est connue par defaut: case vaut vrai, sinon faux
# Noter les transactions sans adresses
df["address_known"] = df["adresse_nom_voie"].apply(
lambda y: True if isinstance(y, str) else False
)
addresses = recover_adress(df, year) # get addresses missing
# addresses = pickle.load(open("address_2020.pickle", "rb"))
add_addresses_to_df(df, addresses) # add adresses to dataframe
# nom de voie de l'adresse en format titre : Une Majuscule A Chaque Mot
df["adresse_nom_voie"] = (
df["adresse_nom_voie"].astype(str).apply(lambda y: y.title())
)
df["code_postal"] = df["code_postal"].apply(lambda y: format_code_postal(y))
# Numero d'adresse converti en entier pour enlever les virgule puis en texte pour le combiner avec le nom de voie
df["adresse_numero"] = df["adresse_numero"].apply(
lambda x: format_adresse_number(x)
)
# Nombre de pieces principale -> nombre entier pour enlever virgule
df["nombre_pieces_principales"] = df["nombre_pieces_principales"].astype("Int32")
df["valeur_fonciere"] = df["valeur_fonciere"].apply(
lambda y: format_valeur_fonciere(y)
)
# Créer une nouvelle colonne qui contient l'adresse entière
df["adresse"] = (
df["adresse_numero"]
+ " "
+ df["adresse_nom_voie"]
+ " "
+ df["code_postal"]
+ " "
+ df["nom_commune"]
).str.strip()
# convertie les dates en format lise par humain et ordinateur
df["date_mutation"] = pd.to_datetime(df["date_mutation"])
# premiere version du tableau nettoyé
return df
if __name__ == "__main__":
all_years = [
2014,
2015,
2016,
2017,
2018,
2019,
] # 2014, 2015, 2016, 2017, 2018, 2019, 2020
for year in all_years:
df = pd.read_csv(
f"/home/vee/Documents/MathieuMemoire/dvf/raw_data/data_dvf_{year}.csv",
low_memory=False,
)
df = clean_dataframe_DVF(df)
df.to_csv(
f"/home/vee/Documents/MathieuMemoire/dvf/formatted_data/data_dvf_{year}.csv"
)
print("DONE ", year)