pkaf

import pandas as pd
import numpy as np

def describe(x):
    data = [np.mean(x), x.std(),  x.median() - x.quantile(0.25), 
            x.median(), x.quantile(0.75) - x.median(), np.sum(x)]
    names = ['mean', 'std', '-err', '50%', '+err', 'sum']
    return pd.DataFrame(data, index=names).T
  
#functional call

pkaf

pkaf

pkaf

import numpy as np

def weighted_median(x, weights):
    
    #Sorting item and weights first
    sort    = np.argsort(x)
    x       = x[sort]
    weights = weights[sort]
    
    #Less than cumulative frequencies

pkaf

import scipy.stats as st
import numpy as np

def err_pdf( loc, scale, size=15, cutoff=5):
    """
    PURPOSE
    --------
    Returns gaussian pdf at grid x that can be used as an error distribution.
    Here, grid x is linearly spaced between -cutoff to +cutoff sigma of the loc.
    

pkaf

pkaf