如何将 statsmodels 测试的结果导出到 CSV？

Question

I am just learning how to use the statsmodels (Python module) to perform such things as regression analysis, testing for normality and homogenous variance. I am working with several data sets (usually as CSV files) and would like to write a script to help me do this more efficiently. My data is just a set of numbers.

Example of data

column1, column2
2.80609,2.80609
2.39059,1.6697286666666666
3.6487540000000003,1.8243770000000001
1.8582885714285717,3.0046419047619044
2.587834,1.7252226666666666
...

Specifically, I would like to:

(1) Iterate a test over several files

(2) Save the result of each test to a new row in a new file

An example of one of the tests I am using.

data = pd.read_csv("data.csv") 

column1 = data['column1']
p_value = sm.stats.diagnostic.lilliefors(data, dist='norm', pvalmethod='approx')

The float it exports look like this.

(0.08557045418097009, 7.144631930303909e-50)

My adventures into if/else and boolean values resulted in this code that prints a text that would be nifty to have exported as well.

p_value = sm.stats.diagnostic.lilliefors(curvature_length, dist='norm', pvalmethod='approx')[1]
if p_value<0.05:
    print("Data is not normal distributed")
else:
    print("Data is normal distributed")
print(p_value)

Any tips & feedback on how to go about this would be greatly appreciated!

Answer 1

To write the two floats to CSV https://docs.python.org/3/library/csv.html

import csv
with open('eggs.csv', 'w', newline='') as csvfile:
    csvwriter = csv.writer(csvfile)

cycle through the pairs of floats doing

        csvwriter.writerow([float1, float2])