import numpy as np
import pandas as pd
%matplotlib inline

from sklearn.preprocessing import StandardScaler

from statsmodels.tsa.stattools import adfuller
from statsmodels.tsa.seasonal import seasonal_decompose

import time
from tqdm import tqdm 
from scipy  import stats

from sklearn.ensemble import ExtraTreesClassifier

df = pd.read_csv("shortened_sensor.csv")

print('here')

pd.set_option("display.max_rows", None, "display.max_columns", None)

df.head()

#df = df.head(5)

#df.to_csv("shortened_sensor.csv", index = False)

#df = df.head(5)

# Find Duplicate Values
# Results will be the list of duplicate values
# If no duplicate values, nothing will list.

df[df['timestamp'].duplicated(keep=False)]

df.isnull().sum()

df['machine_status'].value_counts()

# Convert timestamp column into data type into datetime

df['timestamp'] = pd.to_datetime(df['timestamp'])

# Create a Series
time_period = pd.Series([])

# Assign values to series
for i in tqdm(range(df.shape[0])):
    if (df["timestamp"][i].hour >= 4) and (df["timestamp"][i].hour < 10):
        time_period[i]="Morning"  
    elif (df["timestamp"][i].hour >= 10) and (df["timestamp"][i].hour < 16):
        time_period[i]="Noon"  
    elif (df["timestamp"][i].hour >= 16) and (df["timestamp"][i].hour < 22):   
        time_period[i]="Evening"
    else:    
        time_period[i]="Night"

# Insert new column time period
df.insert(2, 'time_period', time_period)

# The columns sensor_00, sensor_06, sensor-07, sensor_08, sensor_09, sensor-51
# Missing values are filled with median value of respective columns

df['sensor_00'].fillna(df['sensor_00'].median(), inplace=True)
df['sensor_06'].fillna(df['sensor_06'].median(), inplace=True)
df['sensor_07'].fillna(df['sensor_07'].median(), inplace=True)
df['sensor_08'].fillna(df['sensor_08'].median(), inplace=True)
df['sensor_09'].fillna(df['sensor_09'].median(), inplace=True)
df['sensor_51'].fillna(df['sensor_51'].median(), inplace=True)

df['sensor_01'].fillna(df['sensor_01'].median(), inplace=True)
df['sensor_02'].fillna(df['sensor_02'].median(), inplace=True)
df['sensor_03'].fillna(df['sensor_03'].median(), inplace=True)
df['sensor_04'].fillna(df['sensor_04'].median(), inplace=True)
df['sensor_05'].fillna(df['sensor_05'].median(), inplace=True)
df['sensor_10'].fillna(df['sensor_10'].median(), inplace=True)

df

df1 = df.copy()

df.drop(["Unnamed: 0","timestamp","time_period","machine_status"], axis = 1, inplace=True)

df.head()

scaler=StandardScaler()
df=scaler.fit_transform(df)

df

columns = [f'sensor_{idx:02d}' for idx in range(52)]
df2 = pd.DataFrame(df, columns=columns)



df2["machine_status"] = df1["machine_status"]

df2.head()

# Separating the dependent and independent varibale

y = df2 [ 'machine_status' ]
X = df2.drop(['machine_status', 'sensor_15'], axis = 1 )




X.head()

model = ExtraTreesClassifier()
model.fit(X, y)
print(model.feature_importances_)

Error:
C:\Users\james\AppData\Local\Temp\ipykernel_11652\2032974972.py:45: FutureWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.
  time_period = pd.Series([])
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 5/5 [00:00<00:00, 393.24it/s]
[0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0.]