Universal-Stat-Analysis

Limbitless Universal Statistical Analysis Script

NORMALIZATION SCRIPT

• Load in desired .xslx or .csv sheet into ‘excel_sheets’ folder

• Normalize data by running the command in terminal: python normalize.py excel_sheets/insertname.xlsx (or .csv)

• This loads a new .csv file inside the excel_sheets file which will be read by the stats_runner.py script to perform statistical analysis

• Run desired test in terminal based on the command references below inside stat_runner.py script

STAT RUNNER SCRIPT

---------- Independent t-test ----------

FULL COHORT (Cohort 1 vs 2)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_ind --value value --group Cohort --levels 1,2

COHORT 1 vs 2 (Females only)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_ind --value value --group Cohort --levels 1,2 --filter Sex=F

EQUAL VARIANCES (Student t-test)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_ind --equal_var --value value --group Cohort --levels 1,2

AGE RANGE (18–20)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_ind --value value --group Cohort --levels 1,2 --filter "Age>=18,Age<=20"

---------- Paired t-test ----------

R vs L (needs SubjectID)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_rel --value value --subject SubjectID --condition side --levels R,L

R vs L (Cohort 1 only)

python stats_runner.py excel_sheets/insertname_tidy.csv --ttest_rel --value value --subject SubjectID --condition side --levels R,L --filter Cohort=1

---------- One-way ANOVA ----------

Across trial_id (all cohorts)

python stats_runner.py excel_sheets/insertname_tidy.csv --anova1 --value value --factor trial_id

Across trial_id (Females only)

python stats_runner.py excel_sheets/insertname_tidy.csv --anova1 --value value --factor trial_id --filter Sex=F

Across trial_id (Age 18–20)

python stats_runner.py excel_sheets/insertname_tidy.csv --anova1 --value value --factor trial_id --filter "Age>=18,Age<=20"

---------- Two-way ANOVA ----------

Cohort × side

python stats_runner.py excel_sheets/insertname_tidy.csv --anova2 --value value --factorA Cohort --factorB side

Cohort × side (Females only)

python stats_runner.py excel_sheets/insertname_tidy.csv --anova2 --value value --factorA Cohort --factorB side --filter Sex=F

---------- Correlations ----------

Pearson (FULL COHORT)

python stats_runner.py excel_sheets/insertname_tidy.csv --pearson --x Sleep_hours --y value

Spearman (COHORT 2 only)

python stats_runner.py excel_sheets/insertname_tidy.csv --spearman --x Screen_time --y value --filter Cohort=2

Kendall (BY GENDER: Females)

python stats_runner.py excel_sheets/insertname_tidy.csv --kendall --x Age --y value --filter Sex=F

Pearson (BY AGE 18–20)

python stats_runner.py excel_sheets/insertname_tidy.csv --pearson --x Sleep_hours --y value --filter "Age>=18,Age<=20"

---------- Tukey HSD ----------

Pairwise comparisons across trial_id

python stats_runner.py excel_sheets/insertname_tidy.csv --tukey --value value --group trial_id

Tukey across trial_id (Cohort 1 only)

python stats_runner.py excel_sheets/insertname_tidy.csv --tukey --value value --group trial_id --filter Cohort=1

---------- FDR Adjustment ----------

Adjust p-values in a separate CSV

python stats_runner.py excel_sheets/pvals.csv --fdr --pcol pval --alpha 0.05

---------- Bonus combos ----------

Pearson (Cohort 1, Females, Age 18–20)

python stats_runner.py excel_sheets/insertname_tidy.csv --pearson --x Sleep_hours --y value --filter "Cohort=1,Sex=F,Age>=18,Age<=20"

Two-way ANOVA (Cohort × side, Age > 18)

python stats_runner.py excel_sheets/insertname_tidy.csv --anova2 --value value --factorA Cohort --factorB side --filter "Age>18"