Issue
I am following the process shown on Wine Quality Prediction End-to-End ML Project on Krish Naik's YouTube channel to do a Flight Fare Prediction Project.
I run this cell of data transformation pipeline on 03_data_transformation.ipynb:
try:
config = ConfigurationManager()
data_transformation_config = config.get_data_transformation_config()
data_transformation = DataTransformation(config=data_transformation_config)
# data_transformation.train_test_spliting()
# New Line
data_transformation.initiate_data_transformation()
except Exception as e:
raise e
I get this error:
AttributeError: 'DataTransformation' object has no attribute 'data_transformation_config'
Here is the traceback:
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
g:\Machine_Learning_Projects\iNeuron internship\Flight-Fare-Prediction-End-to-End-ML-Project\research\03_data_transformation.ipynb Cell 10 line 9
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=6'>7</a> data_transformation.initiate_data_transformation()
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=7'>8</a> except Exception as e:
----> <a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=8'>9</a> raise e
g:\Machine_Learning_Projects\iNeuron internship\Flight-Fare-Prediction-End-to-End-ML-Project\research\03_data_transformation.ipynb Cell 10 line 7
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=3'>4</a> data_transformation = DataTransformation(config=data_transformation_config)
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=4'>5</a> # data_transformation.train_test_spliting()
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=5'>6</a> # New Line
----> <a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=6'>7</a> data_transformation.initiate_data_transformation()
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=7'>8</a> except Exception as e:
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=8'>9</a> raise e
g:\Machine_Learning_Projects\iNeuron internship\Flight-Fare-Prediction-End-to-End-ML-Project\research\03_data_transformation.ipynb Cell 10 line 8
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=77'>78</a> logger.info(f' transformed df data head: \n{df.head().to_string()}')
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=79'>80</a> # df.to_csv(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=80'>81</a> # New Line
---> <a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=81'>82</a> df.to_excel(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=82'>83</a> logger.info("transformed data is stored")
<a href='vscode-notebook-cell:/g%3A/Machine_Learning_Projects/iNeuron%20internship/Flight-Fare-Prediction-End-to-End-ML-Project/research/03_data_transformation.ipynb#X12sZmlsZQ%3D%3D?line=83'>84</a> df.head(1)
AttributeError: 'DataTransformation' object has no attribute 'data_transformation_config'
Here is the code of data transformation cell, which contains data_transformation_config:
class DataTransformation:
# New Function Added
# https://github.com/yash1314/Flight-Price-Prediction/blob/main/src/utils.py
def convert_to_minutes(self, duration):
try:
hours, minute = 0, 0
for i in duration.split():
if 'h' in i:
hours = int(i[:-1])
elif 'm' in i:
minute = int(i[:-1])
return hours * 60 + minute
except :
return None
def __init__(self, config: DataTransformationConfig):
self.config = config
## Note: You can add different data transformation techniques such as Scaler, PCA and all
#You can perform all kinds of EDA in ML cycle here before passing this data to the model
# I am only adding train_test_spliting cz this data is already cleaned up
# New Code Added Start
def initiate_data_transformation(self):
## reading the data
# df = pd.read_csv(self.config.data_path)
# New Line
df = pd.read_excel(self.config.data_path)
logger.info('Read data completed')
logger.info(f'df dataframe head: \n{df.head().to_string()}')
## dropping null values
df.dropna(inplace = True)
## Date of journey column transformation
df['journey_date'] = pd.to_datetime(df['Date_of_Journey'], format ="%d/%m/%Y").dt.day
df['journey_month'] = pd.to_datetime(df['Date_of_Journey'], format ="%d/%m/%Y").dt.month
## encoding total stops.
df.replace({'Total_Stops': {'non-stop' : 0, '1 stop': 1, '2 stops': 2, '3 stops': 3, '4 stops': 4}}, inplace = True)
## ecoding airline, source, and destination
df_airline = pd.get_dummies(df['Airline'], dtype=int)
df_source = pd.get_dummies(df['Source'], dtype=int)
df_dest = pd.get_dummies(df['Destination'], dtype=int)
## dropping first columns of each categorical variables.
df_airline.drop('Trujet', axis = 1, inplace = True)
df_source.drop('Banglore', axis = 1, inplace = True)
df_dest.drop('Banglore', axis = 1, inplace = True)
df = pd.concat([df, df_airline, df_source, df_dest], axis = 1)
## handling duration column
# df['duration'] = df['Duration'].apply(convert_to_minutes)
# New Line Added
df['duration'] = df['Duration'].apply(self.convert_to_minutes)
upper_time_limit = df.duration.mean() + 1.5 * df.duration.std()
df['duration'] = df['duration'].clip(upper = upper_time_limit)
## encodign duration column
bins = [0, 120, 360, 1440] # custom bin intervals for 'Short,' 'Medium,' and 'Long'
labels = ['Short', 'Medium', 'Long'] # creating labels for encoding
df['duration'] = pd.cut(df['duration'], bins=bins, labels=labels)
df.replace({'duration': {'Short':1, 'Medium':2, 'Long': 3}}, inplace = True)
## dropping the columns
cols_to_drop = cols_to_drop = ['Airline', 'Date_of_Journey', 'Source', 'Destination', 'Route', 'Dep_Time', 'Arrival_Time', 'Duration', 'Additional_Info', 'Delhi', 'Kolkata']
df.drop(cols_to_drop, axis = 1, inplace = True)
logger.info('df data transformation completed')
logger.info(f' transformed df data head: \n{df.head().to_string()}')
# df.to_csv(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
# New Line
df.to_excel(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
logger.info("transformed data is stored")
df.head(1)
## splitting the data into training and target data
X = df.drop('Price', axis = 1)
y = df['Price']
## accessing the feature importance.
select = ExtraTreesRegressor()
select.fit(X, y)
# plt.figure(figsize=(12, 8))
# fig_importances = pd.Series(select.feature_importances_, index=X.columns)
# fig_importances.nlargest(20).plot(kind='barh')
# ## specify the path to the "visuals" folder using os.path.join
# visuals_folder = 'visuals'
# if not os.path.exists(visuals_folder):
# os.makedirs(visuals_folder)
# ## save the plot in the visuals folder
# plt.savefig(os.path.join(visuals_folder, 'feature_importance_plot.png'))
# logger.info('feature imp figure saving is successful')
## further Splitting the data.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 42, shuffle = True)
logger.info('final splitting the data is successful')
## returning splitted data and data_path.
return (
X_train,
X_test,
y_train,
y_test,
self.data_transformation_config.transformed_data_file_path
)class DataTransformation:
# New Function Added
# https://github.com/yash1314/Flight-Price-Prediction/blob/main/src/utils.py
def convert_to_minutes(self, duration):
try:
hours, minute = 0, 0
for i in duration.split():
if 'h' in i:
hours = int(i[:-1])
elif 'm' in i:
minute = int(i[:-1])
return hours * 60 + minute
except :
return None
def __init__(self, config: DataTransformationConfig):
self.config = config
## Note: You can add different data transformation techniques such as Scaler, PCA and all
#You can perform all kinds of EDA in ML cycle here before passing this data to the model
# I am only adding train_test_spliting cz this data is already cleaned up
# New Code Added Start
def initiate_data_transformation(self):
## reading the data
# df = pd.read_csv(self.config.data_path)
# New Line
df = pd.read_excel(self.config.data_path)
logger.info('Read data completed')
logger.info(f'df dataframe head: \n{df.head().to_string()}')
## dropping null values
df.dropna(inplace = True)
## Date of journey column transformation
df['journey_date'] = pd.to_datetime(df['Date_of_Journey'], format ="%d/%m/%Y").dt.day
df['journey_month'] = pd.to_datetime(df['Date_of_Journey'], format ="%d/%m/%Y").dt.month
## encoding total stops.
df.replace({'Total_Stops': {'non-stop' : 0, '1 stop': 1, '2 stops': 2, '3 stops': 3, '4 stops': 4}}, inplace = True)
## ecoding airline, source, and destination
df_airline = pd.get_dummies(df['Airline'], dtype=int)
df_source = pd.get_dummies(df['Source'], dtype=int)
df_dest = pd.get_dummies(df['Destination'], dtype=int)
## dropping first columns of each categorical variables.
df_airline.drop('Trujet', axis = 1, inplace = True)
df_source.drop('Banglore', axis = 1, inplace = True)
df_dest.drop('Banglore', axis = 1, inplace = True)
df = pd.concat([df, df_airline, df_source, df_dest], axis = 1)
## handling duration column
# df['duration'] = df['Duration'].apply(convert_to_minutes)
# New Line Added
df['duration'] = df['Duration'].apply(self.convert_to_minutes)
upper_time_limit = df.duration.mean() + 1.5 * df.duration.std()
df['duration'] = df['duration'].clip(upper = upper_time_limit)
## encodign duration column
bins = [0, 120, 360, 1440] # custom bin intervals for 'Short,' 'Medium,' and 'Long'
labels = ['Short', 'Medium', 'Long'] # creating labels for encoding
df['duration'] = pd.cut(df['duration'], bins=bins, labels=labels)
df.replace({'duration': {'Short':1, 'Medium':2, 'Long': 3}}, inplace = True)
## dropping the columns
cols_to_drop = cols_to_drop = ['Airline', 'Date_of_Journey', 'Source', 'Destination', 'Route', 'Dep_Time', 'Arrival_Time', 'Duration', 'Additional_Info', 'Delhi', 'Kolkata']
df.drop(cols_to_drop, axis = 1, inplace = True)
logger.info('df data transformation completed')
logger.info(f' transformed df data head: \n{df.head().to_string()}')
# df.to_csv(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
# New Line
df.to_excel(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
logger.info("transformed data is stored")
df.head(1)
## splitting the data into training and target data
X = df.drop('Price', axis = 1)
y = df['Price']
## accessing the feature importance.
select = ExtraTreesRegressor()
select.fit(X, y)
# plt.figure(figsize=(12, 8))
# fig_importances = pd.Series(select.feature_importances_, index=X.columns)
# fig_importances.nlargest(20).plot(kind='barh')
# ## specify the path to the "visuals" folder using os.path.join
# visuals_folder = 'visuals'
# if not os.path.exists(visuals_folder):
# os.makedirs(visuals_folder)
# ## save the plot in the visuals folder
# plt.savefig(os.path.join(visuals_folder, 'feature_importance_plot.png'))
# logger.info('feature imp figure saving is successful')
## further Splitting the data.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 42, shuffle = True)
logger.info('final splitting the data is successful')
## returning splitted data and data_path.
return (
X_train,
X_test,
y_train,
y_test,
self.data_transformation_config.transformed_data_file_path
)
Here is the code of configuration manager, which contains get_data_transformation_config() function.
class ConfigurationManager:
def __init__(
self,
config_filepath = CONFIG_FILE_PATH,
params_filepath = PARAMS_FILE_PATH,
schema_filepath = SCHEMA_FILE_PATH):
self.config = read_yaml(config_filepath)
self.params = read_yaml(params_filepath)
self.schema = read_yaml(schema_filepath)
create_directories([self.config.artifacts_root])
def get_data_transformation_config(self) -> DataTransformationConfig:
config = self.config.data_transformation
create_directories([config.root_dir])
data_transformation_config = DataTransformationConfig(
root_dir=config.root_dir,
data_path=config.data_path,
)
return data_transformation_config
Here is my file in GitHub.
My file encoding is UTF-8.
Would you please help me to fix this issue?
Solution
The issue was using the wrong variable.
Instead of
df.to_excel(self.data_transformation_config.transformed_data_file_path, index = False, header= True)
It really should be:
df.to_excel(self.config.transformed_data_file_path, index = False, header= True)
and change the reference to this line:
return (
X_train,
X_test,
y_train,
y_test,
self.data_transformation_config.transformed_data_file_path
)
to
return (
X_train,
X_test,
y_train,
y_test,
self.config.transformed_data_file_path
)
Answered By - ewokx
0 comments:
Post a Comment
Note: Only a member of this blog may post a comment.