### Install pycombat

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Install the pycombat package using pip.

```bash
pip install pycombat
```

--------------------------------

### Initialize Combat instance

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Create an instance of the Combat class. Optional parameters include 'method' ('p' for parametric) and 'conv' for convergence criterion.

```python
combat = Combat()
```

--------------------------------

### Import necessary libraries

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Import NumPy and Pandas for data manipulation and sys for system-specific parameters.

```python
import numpy as np
import pandas as pd
import sys
```

--------------------------------

### Combat Class Initialization

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Initializes an instance of the Combat class. Users can specify parameters to control the harmonization process.

```APIDOC
## Combat Class Initialization

### Description
Initializes an instance of the Combat class. Users can specify parameters to control the harmonization process.

### Parameters
- **method** (string) - Optional - Specifies the harmonization method, either 'p' for parametric or 'np' for non-parametric (not implemented yet).
- **conv** (float) - Optional - The criterion to decide when to stop the EB optimization step. Defaults to 0.0001.
```

--------------------------------

### Prepare covariates matrix (X)

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Create the matrix X of effects of interest by combining dummy-encoded categorical variables (e.g., 'cancer') and continuous variables (e.g., 'age'). Ensure categorical features have their first level dropped using pd.get_dummies with drop_first=True.

```python
X = np.column_stack((pd.get_dummies(pheno.cancer.values, drop_first=True).values,
                    pheno.age.values))

print(X[:10,:])
```

--------------------------------

### Import neuroCombat for comparison

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Import the neuroCombat function for comparison purposes.

```python
from neuroCombat import neuroCombat
```

--------------------------------

### Import Combat class

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Import the Combat class from the pycombat library.

```python
from pycombat import Combat
```

--------------------------------

### Fit and Transform Data with Combat

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Combine the fitting and transformation steps into a single call using the 'fit_transform' method.

```python
Y_adjusted = combat.fit_trasnform(Y=Y, b=b, X=X, C=C)
```

--------------------------------

### Combat.fit_transform

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Combines the fit and transform steps into a single operation.

```APIDOC
## Combat.fit_transform

### Description
Combines the fit and transform steps into a single operation.

### Method
fit_transform

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **Y** (matrix) - Required - The matrix of response variables, with dimensions [observations x features].
- **b** (array) - Required - The array of batch labels for each observation.
- **X** (matrix) - Optional - The matrix of effects of interest to keep, with dimensions [observations x features_interest].
- **C** (matrix) - Optional - The matrix of covariates to remove, with dimensions [observations x features_covariates].

### Response
#### Success Response (200)
- **Y_adjusted** (matrix) - The harmonized data matrix.
```

--------------------------------

### Combat.transform

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Adjusts the data using the parameters learned during the fit step.

```APIDOC
## Combat.transform

### Description
Adjusts the data using the parameters learned during the fit step.

### Method
transform

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **Y** (matrix) - Required - The matrix of response variables, with dimensions [observations x features].
- **b** (array) - Required - The array of batch labels for each observation.
- **X** (matrix) - Optional - The matrix of effects of interest to keep, with dimensions [observations x features_interest].
- **C** (matrix) - Optional - The matrix of covariates to remove, with dimensions [observations x features_covariates].

### Response
#### Success Response (200)
- **Y_adjusted** (matrix) - The harmonized data matrix.
```

--------------------------------

### Combat.fit

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Fits the Combat model to the data by finding the optimal parameters for the linear mixed model.

```APIDOC
## Combat.fit

### Description
Fits the Combat model to the data by finding the optimal parameters for the linear mixed model.

### Method
fit

### Parameters
#### Path Parameters
None

#### Query Parameters
None

#### Request Body
- **Y** (matrix) - Required - The matrix of response variables, with dimensions [observations x features].
- **b** (array) - Required - The array of batch labels for each observation.
- **X** (matrix) - Optional - The matrix of effects of interest to keep, with dimensions [observations x features_interest].
- **C** (matrix) - Optional - The matrix of covariates to remove, with dimensions [observations x features_covariates].
```

--------------------------------

### Fit Combat Model

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Fit the Combat model by calling the 'fit' method with the response variables (Y), batch labels (b), effects of interest (X), and covariates to remove (C). Ensure categorical features in X and C are handled to avoid singularity.

```python
combat.fit(Y=Y, b=b, X=X, C=C)
```

--------------------------------

### Transform Data with Combat

Source: https://github.com/coaxlab/pycombat/blob/master/README.md

Adjust the data using the fitted Combat model by calling the 'transform' method with the same input parameters.

```python
Y_adjusted = combat.transform(Y=Y, b=b, X=X, C=C)
```

--------------------------------

### Transform data with Combat

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Apply the learned parameters from the fitted Combat model to adjust the dataset (Y) using the batch labels (b) and covariates (X).

```python
Y_combat = combat.transform(Y=Y, b=b, X=X)
```

--------------------------------

### Load response variables (Y)

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Load the matrix of response variables from a .npy file. This matrix has dimensions [observations x features].

```python
# Y is the matrix of response variables
Y = np.load('bladder-expr.npy')

print("the matrix of response variables has %d observations and %d outcome variables" % (Y.shape[0], Y.shape[1]))
```

--------------------------------

### Fit Combat model

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Fit the Combat model using the response variables (Y), batch labels (b), and the prepared covariates matrix (X). The C parameter for covariates to remove is optional and defaults to None.

```python
combat.fit(Y, b, X=X, C=None) # X and C are None by default, so no need here to write C=None
```

--------------------------------

### Load phenotype data

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Load phenotype data from a tab-delimited text file into a Pandas DataFrame. This includes batch labels and other covariates.

```python
# This loads the set of independent variables, including the batch labels
pheno = pd.read_csv('bladder-pheno.txt', delimiter='	')

pheno.head()
```

--------------------------------

### Compare PyCombat and neuroCombat results

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Calculate the Pearson correlation coefficient between the harmonized data from PyCombat and neuroCombat to verify consistency.

```python
np.corrcoef(Y_combat.flat, Y_neurocombat.flat)
```

--------------------------------

### Extract batch labels

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Extract the batch labels from the phenotype DataFrame. The number of unique batches is then printed.

```python
b = pheno.batch.values

print("We have %d different batches" % len(np.unique(b)))
```

--------------------------------

### Apply neuroCombat harmonization

Source: https://github.com/coaxlab/pycombat/blob/master/test/test.ipynb

Apply harmonization using the neuroCombat function, specifying discrete and continuous covariates, and the batch column.

```python
discrete_cols = ['cancer']
continuous_cols = ['age']
batch_col = 'batch'

Y_neurocombat = neuroCombat(data=Y,
                            covars=pheno,
                            batch_col=batch_col,
                            discrete_cols=discrete_cols,
                            continuous_cols=continuous_cols)
```

=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.