Module stikpetP.other.poho_binomial
Expand source code
import pandas as pd
from ..tests.test_binomial_os import ts_binomial_os
def ph_binomial(data, expCount=None, twoSidedMethod='eqdist', posthoc = "bonferroni"):
'''
Pairwise Binomial Test for Post-Hoc Analysis
--------------------------------------------
This function will perform a one-sample binomial test for each possible pair in the data. It makes use of the ts_binomial_os() function.
Parameters
----------
data : list or pandas series
expCount : pandas dataframe, optional
categories and expected counts
twoSidedMethod : string, optional
method to use for determining two-sided p-value of binomial test. See ts_binomial_os()
posthoc : string, optional
the correction to use, currently only "bonferroni" available
Returns
-------
res : pandas dataframe with:
* *category 1*, the label of the first category
* *category 2*, the label of the second category
* *n1*, the sample size of the first category
* *n2*, the sample size of the second category
* *obs. prop. cat. 1*, the proportion in the sample of the first category
* *exp. prop. cat. 1*, the expected proportion for the first category
* *p-value*, the unadjusted significance
* *adj. p-value*, the adjusted significance
Author
------
Made by P. Stikker
Companion website: https://PeterStatistics.com
YouTube channel: https://www.youtube.com/stikpet
Donations: https://www.patreon.com/bePatron?u=19398076
Examples
--------
>>> pd.set_option('display.width',1000)
>>> pd.set_option('display.max_columns', 1000)
Example 1: pandas series
>>> df1 = pd.read_csv('https://peterstatistics.com/Packages/ExampleData/GSS2012a.csv', sep=',', low_memory=False, storage_options={'User-Agent': 'Mozilla/5.0'})
>>> ex1 = df1['mar1']
>>> ph_binomial(ex1)
category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value
0 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 1.052263e-55
1 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 7.829174e-78
2 MARRIED WIDOWED 972 181 0.843018 0.5 1.407217e-131 1.407217e-130
3 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 1.267980e-195
4 NEVER MARRIED DIVORCED 395 314 0.557123 0.5 3.001933e-03 3.001933e-02
5 NEVER MARRIED WIDOWED 395 181 0.685764 0.5 1.352112e-19 1.352112e-18
6 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 7.075688e-51
7 DIVORCED WIDOWED 314 181 0.634343 0.5 3.295753e-09 3.295753e-08
8 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 1.472395e-33
9 WIDOWED SEPARATED 181 79 0.696154 0.5 2.223544e-10 2.223544e-09
Example 2: pandas series with various settings
>>> ex2 = df1['mar1']
>>> eCounts = pd.DataFrame({'category' : ["MARRIED", "DIVORCED", "NEVER MARRIED", "SEPARATED"], 'count' : [5,5,5,5]})
>>> ph_binomial(ex2, expCount=eCounts)
category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value
0 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 4.697504e-78
1 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 6.313581e-56
2 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 7.607878e-196
3 DIVORCED NEVER MARRIED 314 395 0.442877 0.5 3.001933e-03 1.801160e-02
4 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 8.834367e-34
5 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 4.245413e-51
Example 3: a list
>>> ex3 = ["MARRIED", "DIVORCED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "NEVER MARRIED", "MARRIED", "MARRIED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "MARRIED"]
>>> ph_binomial(ex3)
category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value
0 DIVORCED MARRIED 7 6 0.538462 0.5 1.000000 1
1 DIVORCED NEVER MARRIED 7 4 0.636364 0.5 0.774414 1
2 DIVORCED SEPARATED 7 2 0.777778 0.5 0.343750 1
3 MARRIED NEVER MARRIED 6 4 0.600000 0.5 0.753906 1
4 MARRIED SEPARATED 6 2 0.750000 0.5 0.289062 1
5 NEVER MARRIED SEPARATED 4 2 0.666667 0.5 0.687500 1
'''
if type(data) is list:
data = pd.Series(data)
freq = data.value_counts()
if expCount is None:
#assume all to be equal
n = sum(freq)
k = len(freq)
categories = list(freq.index)
expC = [n/k] * k
else:
#check if categories match
nE = 0
n = 0
for i in range(0, len(expCount)):
nE = nE + expCount.iloc[i,1]
n = n + freq[expCount.iloc[i,0]]
expC = []
for i in range(0,len(expCount)):
expC.append(expCount.iloc[i, 1]/nE*n)
k = len(expC)
categories = list(expCount.iloc[:,0])
res = pd.DataFrame(columns=["category 1", "category 2", "n1", "n2", "obs. prop. cat. 1", "exp. prop. cat. 1", "p-value", "adj. p-value"])
adjFactor = k * (k - 1)/ 2
for i in range(0, k-1):
for j in range(i+1, k):
n1 = freq[categories[i]]
n2 = freq[categories[j]]
obP1 = n1/(n1 + n2)
exP1 = expC[i]/(expC[i]+expC[j])
codes = [categories[i], categories[j]]
sig = ts_binomial_os(data, codes=codes, p0 = exP1)["p-value (2-sided)"][0]
if posthoc == "bonferroni":
adjSig = sig*adjFactor
if adjSig > 1:
adjSig = 1
res.loc[len(res)] = [categories[i], categories[j], n1, n2, obP1, exP1, sig, adjSig]
return resFunctions
def ph_binomial(data, expCount=None, twoSidedMethod='eqdist', posthoc='bonferroni')-
Pairwise Binomial Test for Post-Hoc Analysis
This function will perform a one-sample binomial test for each possible pair in the data. It makes use of the ts_binomial_os() function.
Parameters
data:listorpandas seriesexpCount:pandas dataframe, optional- categories and expected counts
twoSidedMethod:string, optional- method to use for determining two-sided p-value of binomial test. See ts_binomial_os()
posthoc:string, optional- the correction to use, currently only "bonferroni" available
Returns
res:pandas dataframe with:
- category 1, the label of the first category
- category 2, the label of the second category
- n1, the sample size of the first category
- n2, the sample size of the second category
- obs. prop. cat. 1, the proportion in the sample of the first category
- exp. prop. cat. 1, the expected proportion for the first category
- p-value, the unadjusted significance
- adj. p-value, the adjusted significance
Author
Made by P. Stikker
Companion website: https://PeterStatistics.com
YouTube channel: https://www.youtube.com/stikpet
Donations: https://www.patreon.com/bePatron?u=19398076Examples
>>> pd.set_option('display.width',1000) >>> pd.set_option('display.max_columns', 1000)Example 1: pandas series
>>> df1 = pd.read_csv('https://peterstatistics.com/Packages/ExampleData/GSS2012a.csv', sep=',', low_memory=False, storage_options={'User-Agent': 'Mozilla/5.0'}) >>> ex1 = df1['mar1'] >>> ph_binomial(ex1) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 1.052263e-55 1 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 7.829174e-78 2 MARRIED WIDOWED 972 181 0.843018 0.5 1.407217e-131 1.407217e-130 3 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 1.267980e-195 4 NEVER MARRIED DIVORCED 395 314 0.557123 0.5 3.001933e-03 3.001933e-02 5 NEVER MARRIED WIDOWED 395 181 0.685764 0.5 1.352112e-19 1.352112e-18 6 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 7.075688e-51 7 DIVORCED WIDOWED 314 181 0.634343 0.5 3.295753e-09 3.295753e-08 8 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 1.472395e-33 9 WIDOWED SEPARATED 181 79 0.696154 0.5 2.223544e-10 2.223544e-09Example 2: pandas series with various settings
>>> ex2 = df1['mar1'] >>> eCounts = pd.DataFrame({'category' : ["MARRIED", "DIVORCED", "NEVER MARRIED", "SEPARATED"], 'count' : [5,5,5,5]}) >>> ph_binomial(ex2, expCount=eCounts) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 4.697504e-78 1 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 6.313581e-56 2 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 7.607878e-196 3 DIVORCED NEVER MARRIED 314 395 0.442877 0.5 3.001933e-03 1.801160e-02 4 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 8.834367e-34 5 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 4.245413e-51Example 3: a list
>>> ex3 = ["MARRIED", "DIVORCED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "NEVER MARRIED", "MARRIED", "MARRIED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "MARRIED"] >>> ph_binomial(ex3) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 DIVORCED MARRIED 7 6 0.538462 0.5 1.000000 1 1 DIVORCED NEVER MARRIED 7 4 0.636364 0.5 0.774414 1 2 DIVORCED SEPARATED 7 2 0.777778 0.5 0.343750 1 3 MARRIED NEVER MARRIED 6 4 0.600000 0.5 0.753906 1 4 MARRIED SEPARATED 6 2 0.750000 0.5 0.289062 1 5 NEVER MARRIED SEPARATED 4 2 0.666667 0.5 0.687500 1Expand source code
def ph_binomial(data, expCount=None, twoSidedMethod='eqdist', posthoc = "bonferroni"): ''' Pairwise Binomial Test for Post-Hoc Analysis -------------------------------------------- This function will perform a one-sample binomial test for each possible pair in the data. It makes use of the ts_binomial_os() function. Parameters ---------- data : list or pandas series expCount : pandas dataframe, optional categories and expected counts twoSidedMethod : string, optional method to use for determining two-sided p-value of binomial test. See ts_binomial_os() posthoc : string, optional the correction to use, currently only "bonferroni" available Returns ------- res : pandas dataframe with: * *category 1*, the label of the first category * *category 2*, the label of the second category * *n1*, the sample size of the first category * *n2*, the sample size of the second category * *obs. prop. cat. 1*, the proportion in the sample of the first category * *exp. prop. cat. 1*, the expected proportion for the first category * *p-value*, the unadjusted significance * *adj. p-value*, the adjusted significance Author ------ Made by P. Stikker Companion website: https://PeterStatistics.com YouTube channel: https://www.youtube.com/stikpet Donations: https://www.patreon.com/bePatron?u=19398076 Examples -------- >>> pd.set_option('display.width',1000) >>> pd.set_option('display.max_columns', 1000) Example 1: pandas series >>> df1 = pd.read_csv('https://peterstatistics.com/Packages/ExampleData/GSS2012a.csv', sep=',', low_memory=False, storage_options={'User-Agent': 'Mozilla/5.0'}) >>> ex1 = df1['mar1'] >>> ph_binomial(ex1) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 1.052263e-55 1 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 7.829174e-78 2 MARRIED WIDOWED 972 181 0.843018 0.5 1.407217e-131 1.407217e-130 3 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 1.267980e-195 4 NEVER MARRIED DIVORCED 395 314 0.557123 0.5 3.001933e-03 3.001933e-02 5 NEVER MARRIED WIDOWED 395 181 0.685764 0.5 1.352112e-19 1.352112e-18 6 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 7.075688e-51 7 DIVORCED WIDOWED 314 181 0.634343 0.5 3.295753e-09 3.295753e-08 8 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 1.472395e-33 9 WIDOWED SEPARATED 181 79 0.696154 0.5 2.223544e-10 2.223544e-09 Example 2: pandas series with various settings >>> ex2 = df1['mar1'] >>> eCounts = pd.DataFrame({'category' : ["MARRIED", "DIVORCED", "NEVER MARRIED", "SEPARATED"], 'count' : [5,5,5,5]}) >>> ph_binomial(ex2, expCount=eCounts) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 MARRIED DIVORCED 972 314 0.755832 0.5 7.829174e-79 4.697504e-78 1 MARRIED NEVER MARRIED 972 395 0.711046 0.5 1.052263e-56 6.313581e-56 2 MARRIED SEPARATED 972 79 0.924833 0.5 1.267980e-196 7.607878e-196 3 DIVORCED NEVER MARRIED 314 395 0.442877 0.5 3.001933e-03 1.801160e-02 4 DIVORCED SEPARATED 314 79 0.798982 0.5 1.472395e-34 8.834367e-34 5 NEVER MARRIED SEPARATED 395 79 0.833333 0.5 7.075688e-52 4.245413e-51 Example 3: a list >>> ex3 = ["MARRIED", "DIVORCED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "NEVER MARRIED", "MARRIED", "MARRIED", "MARRIED", "SEPARATED", "DIVORCED", "NEVER MARRIED", "NEVER MARRIED", "DIVORCED", "DIVORCED", "MARRIED"] >>> ph_binomial(ex3) category 1 category 2 n1 n2 obs. prop. cat. 1 exp. prop. cat. 1 p-value adj. p-value 0 DIVORCED MARRIED 7 6 0.538462 0.5 1.000000 1 1 DIVORCED NEVER MARRIED 7 4 0.636364 0.5 0.774414 1 2 DIVORCED SEPARATED 7 2 0.777778 0.5 0.343750 1 3 MARRIED NEVER MARRIED 6 4 0.600000 0.5 0.753906 1 4 MARRIED SEPARATED 6 2 0.750000 0.5 0.289062 1 5 NEVER MARRIED SEPARATED 4 2 0.666667 0.5 0.687500 1 ''' if type(data) is list: data = pd.Series(data) freq = data.value_counts() if expCount is None: #assume all to be equal n = sum(freq) k = len(freq) categories = list(freq.index) expC = [n/k] * k else: #check if categories match nE = 0 n = 0 for i in range(0, len(expCount)): nE = nE + expCount.iloc[i,1] n = n + freq[expCount.iloc[i,0]] expC = [] for i in range(0,len(expCount)): expC.append(expCount.iloc[i, 1]/nE*n) k = len(expC) categories = list(expCount.iloc[:,0]) res = pd.DataFrame(columns=["category 1", "category 2", "n1", "n2", "obs. prop. cat. 1", "exp. prop. cat. 1", "p-value", "adj. p-value"]) adjFactor = k * (k - 1)/ 2 for i in range(0, k-1): for j in range(i+1, k): n1 = freq[categories[i]] n2 = freq[categories[j]] obP1 = n1/(n1 + n2) exP1 = expC[i]/(expC[i]+expC[j]) codes = [categories[i], categories[j]] sig = ts_binomial_os(data, codes=codes, p0 = exP1)["p-value (2-sided)"][0] if posthoc == "bonferroni": adjSig = sig*adjFactor if adjSig > 1: adjSig = 1 res.loc[len(res)] = [categories[i], categories[j], n1, n2, obP1, exP1, sig, adjSig] return res