If you wish to make small changes to the codebase, your pull requests are welcome. However, for major changes or ideas on how to improve the library, please create an issue.
MultiTrain is a python module for machine learning, built with the aim of assisting you to find the machine learning model that works best on a particular dataset.
MultiTrain requires:
- matplotlib==3.5.3
- numpy==1.23.3
- pandas==1.4.4
- plotly==5.10.0
- scikit-learn==1.1.2
- xgboost==1.6.2
- catboost==1.0.6
- imbalanced-learn==0.9.1
- seaborn==0.12.0
- lightgbm==3.3.2
- scikit-optimize==0.9.0
Install MultiTrain using:
pip install MultiTrain
If you experience issues or come across a bug while using MultiTrain, make sure to update to the latest version with
pip install --upgrade MultiTrain
If that doesn't fix your bug, create an issue in the issue tracker
The MultiClassifier is a combination of many classifier estimators, each of which is fitted on the training data and returns assessment metrics such as accuracy, balanced accuracy, r2 score, f1 score, precision, recall, roc auc score for each of the models.
#This is a code snippet of how to import the MultiClassifier and the parameters contained in an instance
from MultiTrain import MultiClassifier
train = MultiClassifier(
n_jobs=-1, # Use all available CPU cores
random_state=42, # Ensure reproducibility
max_iter=1000, # Maximum number of iterations for models that require it
custom_models=['LogisticRegression', 'GradientBoostingClassifier'] # If nothing is set here, all available classifiers will be used for training
)This function operates identically like the scikit-learn framework's train test split function. However, it has some extra features. For example, the split method is demonstrated in the code below.
import pandas as pd
from MultiTrain import MultiClassifier
train = MultiClassifier()
df = pd.read_csv("nameofFile.csv")
split = train.split(
data=df,
target="label_column", # Specify the name of the target column here
random_state=42, # Set a random seed
test_size=0.3, # Set the test size to be used for splitting the dataset i.e 0.3 = 70% train, 30% test
auto_cat_encode=True, # Automatically encode all categorical columns
manual_encode={'label': ['cat_feature'], 'onehot': ['city', 'country']}, # Optional manual encoding for select columns (You can't use this with auto_cat_encode)
fix_nan_custom={'column1': 'ffill', 'column2': 'bfill', 'column3': 'interpolate'}, # Specify columns with the strategies to fill with
drop=['unnecessary_column'] # Drop columns that are not needed
)In 'manual_encode', you are expected to pass in the type of encoding you want to perform on the columns in your dataset. The only available encoding types for now are 'label' for label encoding and 'onehot' for one hot encoding.
# Automatic encoding
split = train.split(
data=df,
target='label_column',
test_size=0.2,
auto_cat_encode=True
)
# Label encoding
split = train.split(
data=df,
target='label_column',
test_size=0.2,
manual_encode={'label': ['column1', 'column2']}
)
# Onehot encoding
split = train.split(
data=df,
target='label_column',
test_size=0.2,
manual_encode={'onehot': ['column1', 'column2']}
)
# Label and onehot encoding
split = train.split(
data=df,
target='label_column',
test_size=0.2,
manual_encode={'label': ['column1', 'column2'],
'onehot': ['column3', 'column4']}
)With the help of the 'fix_nan_custom' argument, you may quickly fill in missing values.
You would need to supply a dictionary to the argument in order to fill in the missing values. Each preset key in the dictionary must be used as shown in the example below.
# the three strategies available to fill missing values are ['ffill', 'bfill', 'interpolate']
```python
split = train.split(
data=df,
target='label_column',
test_size=0.2,
fix_nan_custom={'column1': 'ffill', 'column2': 'bfill', 'column3': 'interpolate'}
)Now that the dataset has been split using the split method, it is time to train on it using the fit method. Instead of the standard training in scikit-learn, catboost, or xgboost, this fit method integrates almost all available machine learning algorithms and trains them all on the dataset. It then returns a pandas dataframe including information such as which algorithm is overfitting, which algorithm has the greatest accuracy, and so on. A basic code example for using the fit function is shown below.
import pandas as pd
from MultiTrain import MultiClassifier
train = MultiClassifier()
df = pd.read_csv('file.csv')
split = train.split(data=df
test_size=0.2,
auto_cat_encode=True,
target='label_column'
)
fit = train.fit(
datasplits=split,
sort='accuracy', # The metric to sort the final results
)
# The available metrics to pass into sort are
# 1. accuracy 2. precision 3. recall 4. f1 5. roc_aucNow, we would be looking at the various ways the fit method can be implemented.
import pandas as pd
from sklearn.model_selection import train_test_split
from MultiTrain import MultiClassifier
train = MultiClassifier()
df = pd.read_csv('filename.csv')
X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.2, random_state=42)
datasplits = (X_train, X_test, y_train, y_test)
fit = train.fit(datasplits=datasplit
show_train_score=True, # Only set this to true if you want to compare train equivalent of all the metrics shown on the dataframe
sort='accuracy', # Set a metric here to sort the resulting dataframe by the best performing model based on the metric
custom_metric='log_loss', # If you set a custom metric here, it will be added to the list of metrics displayed on the final table
imbalanced=True, # Only set this to true if you're working with an imbalanced dataset. It adjust metrics calculation for imbalanced data
text=True, # Set this to true if you're working with NLP
vectorizer= 'count', # specify either count or tfidf if you set text to True
pipeline_dict = {'ngram_range': (1, 2), 'encoding': 'utf-8', 'max_features': 5000, 'analyzer': 'word'} # You must pass in a similar dictionary also if you set text to True
return_best_model = 'f1' # If you set this, it will return the single best performing model based on the f1 score metric
) import pandas as pd
from MultiTrain import MultiClassifier
train = MultiClassifier()
df = pd.read_csv('filename.csv')
split = train.split(data=df
test_size=0.2,
auto_cat_encode=True,
target='label_column'
)
fit = train.fit(datasplits=split,
sort='accuracy',
show_train_score=True) import pandas as pd
from MultiTrain import MultiClassifier
train = MultiClassifier()
df = pd.read_csv('filename.csv')
split = train.split(data=df
test_size=0.2,
auto_cat_encode=True,
target='label_column'
)
fit = train.fit(datasplits=split,
sort='accuracy',
show_train_score=True,
text=True,
vectorizer='tfidf',
pipeline_dict = {'ngram_range': (1, 2), 'encoding': 'utf-8', 'max_features': 5000, 'analyzer': 'word'}
) The MultiRegressor is a combination of many classifier estimators, each of which is fitted on the training data and returns assessment metrics for each of the models.
#This is a code snippet of how to import the MultiClassifier and the parameters contained in an instance
from MultiTrain import MultiRegressor
train = MultiRegressor(
n_jobs=-1, # Use all available CPU cores
random_state=42, # Ensure reproducibility
max_iter=1000, # Maximum number of iterations for models that require it
custom_models=['LogisticRegression', 'GradientBoostingClassifier'] # If nothing is set here, all available classifiers will be used for training
)This function operates identically like the scikit-learn framework's train test split function. However, it has some extra features. For example, the split method is demonstrated in the code below.
from MultiTrain import MultiRegressor
train = MultiRegressor()
df = pd.read_csv('sample_data.csv')
split = train.split(data=df
test_size=0.2,
auto_cat_encode=True,
target='label_column'
)If you want to fill missing values using the split function
If you want to encode your categorical columns using the split function
All you need to do is swap out MultiClassifier with MultiRegressor and you're good to go.
Now, we would be looking at the various ways the fit method can be implemented.
import pandas as pd
from sklearn.model_selection import train_test_split
from MultiTrain import MultiRegressor
train = MultiRegressor()
df = pd.read_csv('filename.csv')
X_train, X_test, y_train, y_test = train_test_split(features, labels, test_size=0.2, random_state=42)
datasplits = (X_train, X_test, y_train, y_test)
fit = train.fit(datasplits=datasplit
show_train_score=True, # Only set this to true if you want to compare train equivalent of all the metrics shown on the dataframe
sort='mean_squared_error', # Set a metric here to sort the resulting dataframe by the best performing model based on the metric
custom_metric='r2_score', # If you set a custom metric here, it will be added to the list of metrics displayed on the final table
return_best_model = 'mean_squared_error' # If you set this, it will return the single best performing model based on the mean squared error metric
)
# The metrics available for sorting are
# mean squared error, r2 score, mean absolute error, median absolute error, mean squared log error, explained variance scoreimport pandas as pd
from MultiTrain import MultiRegressor
train = MultiRegressor()
df = pd.read_csv('filename.csv')
split = train.split(data=df
test_size=0.2,
auto_cat_encode=True,
target='label_column'
)
fit = train.fit(datasplits=split,
sort='r2 score',
show_train_score=True)