Module `cognet.util`

Expand source code

from sklearn.decomposition import PCA
import pandas as pd

def assert_None(args,
                raise_error=True):
    '''Make sure args are not None
    '''
    if any(x is None for x in args):
        num_none = sum(x is None for x in args)
        if raise_error:
            string='Nones detected : {}'.format(str(num_none))
            raise ValueError(string)
        else:
            return num_none

def assert_array_dimension(array, dimensions):
    '''Make sure arrays are the right dimensions
    '''
    if len(array.shape) != dimensions:
        raise ValueError('You must pass in a {}-D array!'.format(dimensions))
    
def embed_to_pca(EFILE, OUTFILE):
    """build pca model with embed file
    """
    Ef=pd.read_csv(EFILE,sep=' ',header=None).dropna(axis=1).transpose()
    Ef.columns=['x'+str(i) for i in Ef.columns]
    xf=Ef#.assign(IF=dx.ido)

    pca = PCA(n_components=2).fit(xf)
    ef=pca.fit_transform(xf)
    pd.DataFrame(ef).to_csv(OUTFILE,header=None,index=None)
    return ef

def replace_nan(df):
    """replace nans in a df and fill with empty string
    """
    df = df.replace('nan',np.nan).fillna('')
    return df

Functions

def assert_None(args, raise_error=True)

Make sure args are not None

Expand source code

def assert_None(args,
                raise_error=True):
    '''Make sure args are not None
    '''
    if any(x is None for x in args):
        num_none = sum(x is None for x in args)
        if raise_error:
            string='Nones detected : {}'.format(str(num_none))
            raise ValueError(string)
        else:
            return num_none

def assert_array_dimension(array, dimensions)

Make sure arrays are the right dimensions

Expand source code

def assert_array_dimension(array, dimensions):
    '''Make sure arrays are the right dimensions
    '''
    if len(array.shape) != dimensions:
        raise ValueError('You must pass in a {}-D array!'.format(dimensions))

def embed_to_pca(EFILE, OUTFILE)

build pca model with embed file

Expand source code

def embed_to_pca(EFILE, OUTFILE):
    """build pca model with embed file
    """
    Ef=pd.read_csv(EFILE,sep=' ',header=None).dropna(axis=1).transpose()
    Ef.columns=['x'+str(i) for i in Ef.columns]
    xf=Ef#.assign(IF=dx.ido)

    pca = PCA(n_components=2).fit(xf)
    ef=pca.fit_transform(xf)
    pd.DataFrame(ef).to_csv(OUTFILE,header=None,index=None)
    return ef

def replace_nan(df)

replace nans in a df and fill with empty string

Expand source code

def replace_nan(df):
    """replace nans in a df and fill with empty string
    """
    df = df.replace('nan',np.nan).fillna('')
    return df