-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdataloader.py
141 lines (127 loc) · 6.38 KB
/
dataloader.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
import api
import scraper
import datetime
import operator
import pandas as pd
import cv2
import os
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image
import glob
def data_loader_with_ar(df):
with open("processed_weather_data.txt") as file:
for line in file:
numbers = line.split()
year, month, day, hour, lat, lng, iwv = int(numbers[1]), int(numbers[2]), int(
numbers[3]), int(numbers[4]), float(numbers[5]), float(numbers[6])-360, float(numbers[10])
date = datetime.datetime(year, month, day, hour)
date_string = date.strftime("%Y-%m-%d T%H:%M:%SZ")
pathname = f"images/image_{year}_{month}_{day}_{hour}_{lat}_{lng}.png"
print(lat, lng, iwv)
if os.path.exists(pathname) and df.loc[df["image"] == pathname].shape[0] <= 0:
print("file exists")
else:
continue
scraper.save_image(lat, lng, zoom=8, date=date_string, pathname=pathname)
weather_data = api.get_weather_data(year, month, day, lat, lng)
yield weather_data, pathname, hour
def data_loader_with_no_ar():
while len(os.listdir("no_ar_images")) < 1156:
lat, long = tuple(zip(np.random.uniform(-90., 90., 1),
np.random.uniform(-180., 180., 1)))[0]
year = np.random.randint(2002, 2023)
month = np.random.randint(1, 13)
day = np.random.randint(1, 29)
hour = np.random.randint(0, 24)
date = datetime.datetime(year, month, day, hour)
date_string = date.strftime("%Y-%m-%d T%H:%M:%SZ")
pathname = f"no_ar_images/image_{year}_{month}_{day}_{hour}_{lat}_{long}.png"
scraper.save_image(lat, long, zoom=8,
date=date_string, pathname=pathname)
im = set(
list(map(lambda x: x[0], list(Image.open(pathname).convert("HSV").getdata()))))
if (set(list(Image.open(pathname).getdata())) == {(0, 0, 0, 255)} or len(im.difference({0, 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, 170, 171, 172, 173, 174, 175, 175, 176, 177, 178, 179, 180})) != len(im)):
os.remove(pathname)
continue
weather_data = api.get_weather_data(year, month, day, lat, long)
yield weather_data, pathname, hour
def save_data():
# df = pd.DataFrame(
# columns=["image", "lat", "long", "generationtime_ms", "utc_offset_seconds", "timezone", "elevation", "time", "temperature_2m", "ar", "hour"])
df = pd.read_pickle("data.pkl") if (os.path.exists("data.pkl")) else pd.DataFrame(
columns=["image", "lat", "long", "generationtime_ms", "utc_offset_seconds", "timezone", "elevation", "time", "temperature_2m", "ar", "hour"])
try:
for data, image, hour in data_loader_with_ar(df):
df = df.append({"image": image, "lat": data["latitude"], "long": data["longitude"], "generationtime_ms": data["generationtime_ms"], "utc_offset_seconds": data["utc_offset_seconds"],
"timezone": data["timezone"], "elevation": data["elevation"], "time": data["hourly"]["time"][hour-1], "temperature_2m": data["hourly"]["temperature_2m"][hour-1], "ar": 1, "hour": hour}, ignore_index=True)
except Exception as e:
print(e)
try:
for data, image, hour in data_loader_with_no_ar():
df = df.append({"image": image, "lat": data["latitude"], "long": data["longitude"], "generationtime_ms": data["generationtime_ms"], "utc_offset_seconds": data["utc_offset_seconds"],
"timezone": data["timezone"], "elevation": data["elevation"], "time": data["hourly"]["time"][hour-1], "temperature_2m": data["hourly"]["temperature_2m"][hour-1], "ar": 0, "hour": hour}, ignore_index=True)
except Exception as e:
print(e)
df.to_pickle("data.pkl")
def show_data():
df = pd.read_pickle("data.pkl")
print(df["time"], df["hour"])
print(len(df))
print(df.tail())
print(len(df[df["ar"] == 1]))
print(len(df[df["ar"] == 0]))
print("No Images length", len(os.listdir("no_ar_images/")))
print("Images length", len(os.listdir("images/")))
# df = pd.read_csv("data.csv")
# for index, row in df.iterrows():
# image = cv2.imread(row["image"])
# image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
# plt.imshow(image)
# plt.show()
# with open("weather_data.txt") as file:
# with open("processed_weather_data.txt", "w") as file2:
# lines = []
# for line in file:
# numbers = line.split()
# year, iwv = int(numbers[1]), float(numbers[10])
# if year > 2002 and iwv > 30.0:
# lines.append({"year": year, "iwv": iwv, "line": line})
# lines.sort(key=operator.itemgetter("iwv", "year"))
# for line in lines:
# file2.write(line["line"])
# for weather_data in data_loader():
# print(weather_data)
# save_data()
# df = pd.read_csv("data.csv")
# df = df[df["ar"] == 1]
# df = df.drop(columns=["Unnamed: 0"])
# df.to_csv("data.csv")
# df.to_csv("data.csv")
# show_data()
df = pd.read_pickle("data.pkl")
if os.path.exists("new_data.pkl"):
new_df = pd.read_pickle("new_data.pkl")
else:
new_df = pd.DataFrame(columns=["image", "lat", "long", "temperature",
"humidity", "dewpoint", "precipitation", "ar"])
print(len(new_df))
try:
for i, row in df.iterrows():
if len(new_df) == len(df):
break
if new_df.loc[new_df["image"] == row["image"]].shape[0] > 0:
continue
pathname = row["image"]
if (pathname.split("/")[0] == "no_ar_images"):
pathname = pathname.split("/")[1]
pathname = pathname.split("_")
year, month, day, hour, lat, long = int(pathname[1]), int(pathname[2]), int(
pathname[3]), int(pathname[4]), row["lat"], row["long"]
ar = row["ar"]
weather_data = api.get_weather_data(year, month, day, lat, long)
new_df = new_df.append({"image": row["image"], "lat": lat, "long": long,
"temperature": weather_data["hourly"]["temperature_2m"][hour-1], "humidity": weather_data["hourly"]["relativehumidity_2m"][hour-1], "dewpoint": weather_data["hourly"]["dewpoint_2m"][hour-1], "precipitation": weather_data["hourly"]["precipitation"][hour-1], "ar": ar}, ignore_index=True)
except Exception as e:
print(e)
new_df.to_pickle("new_data.pkl")