diff --git a/__pycache__/t_tests.cpython-314.pyc b/__pycache__/t_tests.cpython-314.pyc new file mode 100644 index 0000000..9b75777 Binary files /dev/null and b/__pycache__/t_tests.cpython-314.pyc differ diff --git a/get_statistics.py b/get_statistics.py index 599da5e..3122377 100644 --- a/get_statistics.py +++ b/get_statistics.py @@ -6,7 +6,7 @@ file = "./data/pnad_trimestral_trimestre_012026.parquet" regions_ufs = { "Sul": [41, 42, 43], "Sudeste": [31, 32, 33, 35], - "Centro-Oeste": [50, 51, 52, 53], + "CentroOeste": [50, 51, 52, 53], "Norte": [11, 12, 13, 14, 15, 16, 17], "Nordeste": [21, 22, 23, 24, 25, 26, 27, 28, 29], } diff --git a/t-tests.py b/t-tests.py deleted file mode 100644 index 6f2e10e..0000000 --- a/t-tests.py +++ /dev/null @@ -1,54 +0,0 @@ -from get_statistics import get_regions, regions_names -import math - -regions = get_regions() - - -def t(n1, mean1, std1, n2, mean2, std2): - return (mean1 - mean2) / math.sqrt( - (math.pow(std1, 2) / n1) + (math.pow(std2, 2) / n2) - ) - - -def run_all_t_tests(): - tests = {} - for name1 in regions_names: - for name2 in regions_names: - if name1 != name2: - tests[f"{name1}_formais_{name2}_informais"] = t( - regions[name1]["formal_count"], - regions[name1]["formal_mean"], - regions[name1]["formal_std"], - regions[name2]["informal_count"], - regions[name2]["informal_mean"], - regions[name2]["informal_std"], - ) - tests[f"{name1}_informais_{name2}_informais"] = t( - regions[name1]["informal_count"], - regions[name1]["informal_mean"], - regions[name1]["informal_std"], - regions[name2]["informal_count"], - regions[name2]["informal_mean"], - regions[name2]["informal_std"], - ) - tests[f"{name1}_informais_{name2}_formais"] = t( - regions[name1]["informal_count"], - regions[name1]["informal_mean"], - regions[name1]["informal_std"], - regions[name2]["formal_count"], - regions[name2]["formal_mean"], - regions[name2]["formal_std"], - ) - tests[f"{name1}_formais_{name2}_formais"] = t( - regions[name1]["formal_count"], - regions[name1]["formal_mean"], - regions[name1]["formal_std"], - regions[name2]["formal_count"], - regions[name2]["formal_mean"], - regions[name2]["formal_std"], - ) - - return tests - - -print(run_all_t_tests()) diff --git a/t_tests.py b/t_tests.py new file mode 100644 index 0000000..645fc10 --- /dev/null +++ b/t_tests.py @@ -0,0 +1,57 @@ +from get_statistics import get_regions, regions_names +import math + +regions = get_regions() + + +def t(n1, mean1, std1, n2, mean2, std2): + return (mean1 - mean2) / math.sqrt( + (math.pow(std1, 2) / n1) + (math.pow(std2, 2) / n2) + ) + + +def run_all_t_tests(): + tests = {} + for name1 in regions_names: + for name2 in regions_names: + region1 = regions[name1] + region2 = regions[name2] + if name1 != name2: + tests[f"{name1}_informal_{name2}_informal"] = t( + region1["informal_count"], + region1["informal_mean"], + region1["informal_std"], + region2["informal_count"], + region2["informal_mean"], + region2["informal_std"], + ) + tests[f"{name1}_formal_{name2}_formal"] = t( + region1["formal_count"], + region1["formal_mean"], + region1["formal_std"], + region2["formal_count"], + region2["formal_mean"], + region2["formal_std"], + ) + tests[f"{name1}_formal_{name2}_informal"] = t( + region1["formal_count"], + region1["formal_mean"], + region1["formal_std"], + region2["informal_count"], + region2["informal_mean"], + region2["informal_std"], + ) + tests[f"{name1}_informal_{name2}_formal"] = t( + region1["informal_count"], + region1["informal_mean"], + region1["informal_std"], + region2["formal_count"], + region2["formal_mean"], + region2["formal_std"], + ) + + return tests + + +if __name__ == "__main__": + print(run_all_t_tests())