Mission_Impossible / Normalizing_Dataset_and_calculating_TrainNumber.py

Upload 5 files

25a0890 verified over 1 year ago

2.4 kB

import pandas as pd

df_raw = pd.read_csv("train_dataset_named.csv") #Reading the CSV file made in the previous code

city_mapping = {"New York": 1, "Boston": 2, "Chicago": 3, "Washington": 4, "Philadelphia": 5}     # Define mappings for normalization
day_mapping = {"Monday": 1, "Tuesday": 2, "Wednesday": 3, "Thursday": 4, "Friday": 5, "Saturday": 6, "Sunday": 7}
train_type_mapping = {"Local": 0, "Express": 1}

def time_to_hours(time_str):      #Converting time string to hours
    hours, minutes = map(int, time_str.split(":"))
    return hours + minutes / 60.0

df_normalized = pd.DataFrame()  #Normalized dataframe

#Normalizing all of the features and converting them into 0-1 Range
df_normalized["departure_station"] = df_raw["departure_station"].map(city_mapping) / 5.0
df_normalized["arrival_station"] = df_raw["arrival_station"].map(city_mapping) / 5.0
df_normalized["departure_time"] = df_raw["departure_time"].apply(time_to_hours) / 24.0
df_normalized["arrival_time"] = df_raw["arrival_time"].apply(time_to_hours) / 24.0
df_normalized["day_of_week"] = df_raw["day_of_week"].map(day_mapping) / 7.0
df_normalized["train_type"] = df_raw["train_type"].map(train_type_mapping)
df_normalized["distance_km"] = df_raw["distance_km"].astype(float) / 1000.0
df_normalized["distance_km"] = (df_normalized["distance_km"] - 0.15) / (1.4 - 0.15)

# Pre calculated weights and biases to calculate train number manually
weights = {
    "departure_station": 0.5,
    "arrival_station": 0.5,
    "departure_time": 0.5,
    "arrival_time": 0.5,
    "day_of_week": 0.5,
    "train_type": 3.5,
    "distance_km": 4.5
}
bias = 0.5

df_normalized["train_number"] = (
    weights["departure_station"] * df_normalized["departure_station"] +
    weights["arrival_station"] * df_normalized["arrival_station"] +
    weights["departure_time"] * df_normalized["departure_time"] +
    weights["arrival_time"] * df_normalized["arrival_time"] +
    weights["day_of_week"] * df_normalized["day_of_week"] +
    weights["train_type"] * df_normalized["train_type"] +
    weights["distance_km"] * df_normalized["distance_km"] +
    bias
).round().astype(int)   #Rounding of train number to integers

df_normalized["train_number"] = df_normalized["train_number"].clip(1, 10)  #Values greater than 10 are clipped to 10

df_normalized.to_csv("normalized_train_dataset_with_train_number.csv", index=False) # Saved as a CSV file

	import pandas as pd

	df_raw = pd.read_csv("train_dataset_named.csv") #Reading the CSV file made in the previous code

	city_mapping = {"New York": 1, "Boston": 2, "Chicago": 3, "Washington": 4, "Philadelphia": 5} # Define mappings for normalization
	day_mapping = {"Monday": 1, "Tuesday": 2, "Wednesday": 3, "Thursday": 4, "Friday": 5, "Saturday": 6, "Sunday": 7}
	train_type_mapping = {"Local": 0, "Express": 1}

	def time_to_hours(time_str): #Converting time string to hours
	hours, minutes = map(int, time_str.split(":"))
	return hours + minutes / 60.0

	df_normalized = pd.DataFrame() #Normalized dataframe

	#Normalizing all of the features and converting them into 0-1 Range
	df_normalized["departure_station"] = df_raw["departure_station"].map(city_mapping) / 5.0
	df_normalized["arrival_station"] = df_raw["arrival_station"].map(city_mapping) / 5.0
	df_normalized["departure_time"] = df_raw["departure_time"].apply(time_to_hours) / 24.0
	df_normalized["arrival_time"] = df_raw["arrival_time"].apply(time_to_hours) / 24.0
	df_normalized["day_of_week"] = df_raw["day_of_week"].map(day_mapping) / 7.0
	df_normalized["train_type"] = df_raw["train_type"].map(train_type_mapping)
	df_normalized["distance_km"] = df_raw["distance_km"].astype(float) / 1000.0
	df_normalized["distance_km"] = (df_normalized["distance_km"] - 0.15) / (1.4 - 0.15)

	# Pre calculated weights and biases to calculate train number manually
	weights = {
	"departure_station": 0.5,
	"arrival_station": 0.5,
	"departure_time": 0.5,
	"arrival_time": 0.5,
	"day_of_week": 0.5,
	"train_type": 3.5,
	"distance_km": 4.5
	}
	bias = 0.5

	df_normalized["train_number"] = (
	weights["departure_station"] * df_normalized["departure_station"] +
	weights["arrival_station"] * df_normalized["arrival_station"] +
	weights["departure_time"] * df_normalized["departure_time"] +
	weights["arrival_time"] * df_normalized["arrival_time"] +
	weights["day_of_week"] * df_normalized["day_of_week"] +
	weights["train_type"] * df_normalized["train_type"] +
	weights["distance_km"] * df_normalized["distance_km"] +
	bias
	).round().astype(int) #Rounding of train number to integers

	df_normalized["train_number"] = df_normalized["train_number"].clip(1, 10) #Values greater than 10 are clipped to 10

	df_normalized.to_csv("normalized_train_dataset_with_train_number.csv", index=False) # Saved as a CSV file