def convert_dataframe(df, to='pandas', return_library=True, dtypes=None):
'''Converts a dataframe to the desired dataframe library format.
Parameters
----------
df : pandas.DataFrame or dask.DataFrame or modin.pandas.DataFrame
Original dataframe which will be converted.
to : string, default 'pandas'
The data library to which format the dataframe will be converted to.
return_library : bool, default True
If set to True, the new dataframe library is also returned as an output.
dtypes : dict, default None
Dictionary that indicates the desired dtype for each column.
e.g. {'Var1': 'float64', 'Var2': 'UInt8', 'Var3': str}
Returns
-------
df : pandas.DataFrame or dask.DataFrame or modin.pandas.dataframe.DataFrame
Converted dataframe, in the desired type.
If return_library == True:
new_pd : pandas or modin.pandas
The dataframe library to which the input dataframe is converted to.
'''
lib = str(to).lower()
if lib == 'pandas':
import pandas as new_pd
elif lib == 'modin':
import modin.pandas as new_pd
else:
raise Exception(
f'ERROR: Currently, convertion to a dataframe of type {to} is not supported. Availabale options are "pandas" and "modin".'
)
converted_df = new_pd.DataFrame(data=df.to_numpy(), columns=df.columns)
du.set_pandas_library(lib)
if dtypes is None:
# Infer adequate dtypes for the dataframe's columns
converted_df = converted_df.infer_objects()
else:
# Set the desired dtypes
converted_df = convert_dtypes(converted_df,
dtypes=dtypes,
inplace=True)
if return_library is True:
return converted_df, new_pd
else:
return converted_df
import numpy as np # Mathematical operations package, allowing also for missing values representation
import torch # PyTorch for tensor and deep learning operations
import plotly.graph_objs as go # Plotly for interactive and pretty plots
import data_utils as du # Data science and machine learning relevant methods
from model_interpreter.model_interpreter import ModelInterpreter # Model interpretability class
import shap # Model-agnostic interpretability package inspired on Shapley values
from datetime import datetime # datetime to use proper date and time formats
import pickle # Save and load Python objects
du.random_seed
du.set_random_seed(42)
du.random_seed
du.set_pandas_library(lib='pandas')
import pixiedust # Debugging in Jupyter Notebook cells
# Change to scripts directory
os.chdir('../../scripts')
import Models # Script with all the machine learning model classes
# Change to parent directory (presumably "eICU-mortality-prediction")
os.chdir('..')
# ## Initializing variables
# Comet ML settings:
import pandas as pd # Pandas to load the data initially
import numpy as np # Mathematical operations package, allowing also for missing values representation
import torch # PyTorch for tensor and deep learning operations
import data_utils as du # Data science and machine learning relevant methods
import os # os handles directory/workspace changes
du.random_seed
du.set_random_seed(42)
du.random_seed
du.use_modin
du.set_pandas_library('pandas')
du.use_modin
import pixiedust # Debugging in Jupyter Notebook cells
# Change to scripts directory
os.chdir('../../scripts')
from Tabular_Dataset import Tabular_Dataset # Dataset class that helps fetching batches of data
import Models # Script with all the machine learning model classes
# Change to parent directory (presumably "eICU-mortality-prediction")
os.chdir('..')
# ## Initializing variables