weekdays = ["M", "Tu", "W", "Th", "F", "Sa", "Su"]
def dow_map(dow):
return weekdays[dow]
tod = ["Night", "Day", "Evening"]
def tod_map(t):
return tod[t]
crimes = decrypt_csv("./data/Playing_Periodicity.csv.enc").drop(
["MonthCreated", "WeekCreated", "DayCreated"], axis=1)
# # fix codes that have turned into dates
crimes.xcor_code = crimes.xcor_code.apply(fix_code)
# create crime description lookup
#codes = crimes.xcor_code.unique()
code_lookup = crimes[["xcor_code", "xcor_lkhoccodename"
]].set_index("xcor_code").drop_duplicates()
# print(code_lookup.head())
# print(len(code_lookup))
# drop descriptions and expand out counts
total = crimes.TotalCreated.sum()
import pandas as pd
from pathlib import Path
from crims.encryption import encrypt_csv, decrypt_csv
if __name__ == "__main__":
data_file = "data/weekly-weights.csv"
assert Path(data_file).is_file(), "Plaintext data file not found"
df = pd.read_csv(data_file)
print(df.head())
encrypt_csv(df, data_file + ".enc", False)
data_file += ".enc"
assert Path(data_file).is_file(), "Encrypted data file not found"
decrypted_df = decrypt_csv(data_file)
print(decrypted_df.head())
# have to check columns separately due to float rounding errors
assert decrypted_df["xcor_code"].equals(df["xcor_code"])
assert decrypted_df["period"].equals(df["period"])
assert decrypted_df["count"].equals(df["count"])
assert decrypted_df["total"].equals(df["total"])
assert (decrypted_df["weight"] - df["weight"]).abs().max() < 1e-15
# 22B = 544
# 30C = 2103
# 8R = 7996
# 105A = 16083
types = ["1", "10A", "17B", "22B", "30C", "8R", "105A"] #[1, 100]
names = pd.read_csv("./data/severity_codes.csv")[[
"description", "code_original"
]]
all_d = pd.read_csv("./daily_adjusted.csv",
index_col=["xcor_code", "TimeWindow"])
all_w = pd.read_csv("./weekly_adjusted.csv",
index_col=["xcor_code", "DayNumber"])
#all_p = pd.read_csv("./period_adjusted.csv", index_col=["xcor_code", "period"])
all_p = decrypt_csv("./data/weekly-weights.csv.enc",
index_col=["xcor_code", "period"])
x = []
for wday in ["M", "Tu", "W", "Th", "F", "Sa", "Su"]:
for shift in ["n", "d", "e"]:
x.append("%s-%s" % (wday, shift))
#print(x)
# xsize = 1200
# ysize = 900
# dpi = 200
#plt.tight_layout()
for t in types:
from crims.encryption import encrypt_csv, decrypt_csv if __name__ == "__main__": data_file="test/crime-sample.csv" assert Path(data_file).is_file(), "Plaintext data file not found" df = pd.read_csv(data_file, index_col="id") print(df.head()) encrypt_csv(df, data_file+".enc") data_file += ".enc" assert Path(data_file).is_file(), "Encrypted data file not found" decrypted_df = decrypt_csv(data_file, index_col="id") print(decrypted_df.head()) # have to check columns separately due to float rounding errors assert decrypted_df.index.equals(df.index) assert decrypted_df["MSOA"].equals(df["MSOA"]) assert decrypted_df["crime_type"].equals(df["crime_type"]) assert decrypted_df["code"].equals(df["code"]) assert decrypted_df["description"].equals(df["description"]) assert decrypted_df["time"].equals(df["time"]) assert decrypted_df["suspect"].equals(df["suspect"]) assert decrypted_df["severity"].equals(df["severity"]) # remove the encrypted file os.remove(data_file)