IndexError: invalid index to scalar variable error - numpy

Question

For each gene, I want to perform McNemar's test and then evaluate if the p-value > 0.05. I want to calculate the number of genes that pass the test.
My code raised IndexError: invalid index to scalar variable.
import pandas as pd
from statsmodels.stats.contingency_tables import mcnemar
from statsmodels.stats.gof import chisquare_effectsize
from statsmodels.stats.power import GofChisquarePower
def generate_gene_df(gene, n):
df = pd.DataFrame.from_dict(
{"Gene" : gene,
"Cells": (f'Cell{x}' for x in range(1, n+1)),
"Control": np.random.choice([1,0], p=[0.1, 0.9], size=n),
"Experimental": np.random.choice([1,0], p=[0.1+0.1, 0.9-0.1], size=n)},
orient='columns'
)
df = df.set_index(["Gene","Cells"])
return df
table = pd.crosstab([df["Gene"], df["Cells"]],
[df["Control"], df["Experimental"]]).to_numpy()
# List of simulated genes
gene_df_list = [generate_gene_df(gene, n) for gene in gene_list]
df = pd.concat(gene_df_list)
df = df.reset_index()
alpha=0.05
p_adjusted=[]
pass_test = []
# McNemar's test
result = mcnemar(table, exact=True)
# Bonferroni correction
p_adjusted = multipletests(result.pvalue, alpha=alpha, method="bonferroni")
for index, value in np.ndenumerate(table):
if result.pvalue > alpha:
np.append(pass_test, result.pvalue[index])
num_passed = len(pass_test)
print("Number of genes that failed to reject H0 is: %i" %num_passed)
Traceback:
---------------------------------------------------------------------------
IndexError Traceback (most recent call last)
/tmp/ipykernel_593/1521754442.py in <module>
11 for index, value in np.ndenumerate(table):
12 if result.pvalue > alpha:
---> 13 np.append(pass_test, result.pvalue[index])
14
15
IndexError: invalid index to scalar variable.

Answer 1

I haven't looked at mcnemar, so don't know what it produces. But the error tells us/youthat result.pvalue is a scalar value, and not indexable.
if result.pvalue > alpha:
np.append(pass_test, result.pvalue[index])
I also can deduce that from the fact that the if line works. It would return an ambiguity error if result.pvalue was an array.
Besides rereading the mcnemar docs, I'd suggest rereading the np.append docs (assuming you even did that!).
Generally we discourage use of np.append in a loop. np.append is not a list append clone