Making a python function that shows status of consumption

Question

I have a project about anomaly detection. I have already identified the predicted values. What I want now is to make the status of the energy consumption based on the predicted values.

These are the basis of the labels:

This is the result that I want:

I have tried the following but I always get an error:

def status(x):
    if df_an['powerconsumption'] <= (df_an['predicted_values'] + (1*df_an['predicted_values'].std())):
        return "Normal"
    elif df_an['powerconsumption'] >= (df_an['predicted_values'] - (1*df_an['predicted_values'].std())):
        return "Normal"
    elif df_an['powerconsumption'] <= (df_an['predicted_values'] + (2*df_an['predicted_values'].std())):
        return "Above Normal"
    elif df_an['powerconsumption'] >= (df_an['predicted_values'] - (2*df_an['predicted_values'].std())):
        return "Above Normal"
    elif df_an['powerconsumption'] <= (df_an['predicted_values'] + (3*df_an['predicted_values'].std())):
        return "Normal"
    elif df_an['powerconsumption'] >= (df_an['predicted_values'] - (3*df_an['predicted_values'].std())):
        return "Normal"
    else:
        return "Anomalous"

for col in df_an.columns:
    df_an['status'] = df_an['powerconsumption'].apply(lambda x: status(x))

I have also tried this code:

filter_method = lambda x: 
'Normal' if (df_an['powerconsumption'] =< (df_an['predicted_values'].mean() + (1*df_an['predicted_values'].std())))
else 'Normal' if df_an['powerconsumption'] => df_an['predicted_values'].mean() - (1*df_an['predicted_values'].std())
else 'Above Normal' if df_an['powerconsumption'] =< df_an['predicted_values'].mean() + (2*df_an['predicted_values'].std())
else 'Above Normal' if df_an['powerconsumption'] => df_an['predicted_values'].mean() - (2*df_an['predicted_values'].std())
else 'Nearing Anomalous' if df_an['powerconsumption'] =< df_an['predicted_values'].mean() + (3*df_an['predicted_values'].std())
else 'Nearing Anomalous' if df_an['powerconsumption'] => df_an['predicted_values'].mean() - (3*df_an['predicted_values'].std())
else 'Anomalous'df_an['powerconsumption'] > df_an['predicted_values'].mean() + (3*df_an['predicted_values'].std())
else 'Anomalous'df_an['powerconsumption'] < df_an['predicted_values'].mean() - (3*df_an['predicted_values'].std())  

I always get a syntax error.

For the first one, this is the error message:

    ---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
C:\Users\Public\Documents\Wondershare\CreatorTemp/ipykernel_4112/2640136280.py in <module>
     16 
     17 for col in df_an.columns:
---> 18     df_an['status'] = df_an['powerconsumption'].apply(lambda x: status(x))

c:\users\lenovo\anaconda3\envs\project\lib\site-packages\pandas\core\series.py in apply(self, func, convert_dtype, args, **kwargs)
   4355         dtype: float64
   4356         """
-> 4357         return SeriesApply(self, func, convert_dtype, args, kwargs).apply()
   4358 
   4359     def _reduce(

c:\users\lenovo\anaconda3\envs\project\lib\site-packages\pandas\core\apply.py in apply(self)
   1041             return self.apply_str()
   1042 
-> 1043         return self.apply_standard()
   1044 
   1045     def agg(self):

c:\users\lenovo\anaconda3\envs\project\lib\site-packages\pandas\core\apply.py in apply_standard(self)
   1100                     values,
   1101                     f,  # type: ignore[arg-type]
-> 1102                     convert=self.convert_dtype,
   1103                 )
   1104 

c:\users\lenovo\anaconda3\envs\project\lib\site-packages\pandas\_libs\lib.pyx in pandas._libs.lib.map_infer()

C:\Users\Public\Documents\Wondershare\CreatorTemp/ipykernel_4112/2640136280.py in <lambda>(x)
     16 
     17 for col in df_an.columns:
---> 18     df_an['status'] = df_an['powerconsumption'].apply(lambda x: status(x))

C:\Users\Public\Documents\Wondershare\CreatorTemp/ipykernel_4112/2640136280.py in status(x)
      1 def status(x):
----> 2     if df_an['powerconsumption'] <= (df_an['predicted_values'] + (1*df_an['predicted_values'].std())):
      3         return "Normal"
      4     elif df_an['powerconsumption'] >= (df_an['predicted_values'] - (1*df_an['predicted_values'].std())):
      5         return "Normal"

c:\users\lenovo\anaconda3\envs\project\lib\site-packages\pandas\core\generic.py in __nonzero__(self)
   1536     def __nonzero__(self):
   1537         raise ValueError(
-> 1538             f"The truth value of a {type(self).__name__} is ambiguous. "
   1539             "Use a.empty, a.bool(), a.item(), a.any() or a.all()."
   1540         )

ValueError: The truth value of a Series is ambiguous. Use a.empty, a.bool(), a.item(), a.any() or a.all().

For the second one, this is the error message:

        File "C:\Users\Public\Documents\Wondershare\CreatorTemp/ipykernel_4112/3716694249.py", line 1
    filter_method = lambda x: 'Normal' if (df_an['powerconsumption'] > (df_an['predicted_values'].mean() + (1*df_an['predicted_values'].std())))
                                                                                                                                                ^
SyntaxError: invalid syntax

Answer 1

You can combine arithmetic calculation of the absolute difference to the mean relative to the standard deviation and pandas.cut:

df = pd.DataFrame({'power': [1000, 1002, 1005, 1009, 1015]+[1000]*20})

df['status'] = pd.cut(df['power'].sub(df['power'].mean().abs())
                                 .div(df['power'].std()),
                      bins=[0, 1, 2, 3, np.inf],
                      labels=['Normal', 'Above Normal',
                              'Nearing Anomalous', 'Anomalous'])

Output:

    power             status
0    1000             Normal
1    1002             Normal
2    1005       Above Normal
3    1009  Nearing Anomalous
4    1015          Anomalous
5    1000             Normal
6    1000             Normal
7    1000             Normal
...