How to use the pyreadstat.read_sas7bdat function in pyreadstat

def test_sas7bdat(self):

        df, meta = pyreadstat.read_sas7bdat(os.path.join(self.basic_data_folder, "sample.sas7bdat"))
        self.assertTrue(df.equals(self.df_pandas))
        self.assertTrue(meta.number_columns == len(self.df_pandas.columns))
        self.assertTrue(meta.number_rows == len(self.df_pandas))

df2["date"] = pd.to_datetime(df2["date"])
df2["dtime"] = pd.to_datetime(df2["dtime"])
df2["time"] = pd.to_datetime(df2["time"])
df2["time"] = df2["time"].apply(lambda x: x.time())

df3 = df2.copy()
df3["date"] = df3["date"].apply(lambda x: x.date())
df3["time"] = df3["time"].apply(lambda x: x.time())


import pyreadstat
sas_file = os.path.join(basic_data_folder, "dates.sas7bdat")
df_sas,meta = pyreadstat.read_sas7bdat(sas_file)
df_sas.equals(df3)

df_sas,meta = pyreadstat.read_sas7bdat(sas_file, dates_as_pandas_datetime=True)
df_sas.equals(df2)


for col in df2.columns:
    if not df2[col].equals(df_sas[col]):
        for indx, (pa, sas) in enumerate(zip(df2[col], df_sas[col])):
            if pa != sas:
                print(indx, pa, sas)

unformatted_csv = os.path.join(self.missing_data_folder, "missing_unformatted.csv")
        formatted_csv = os.path.join(self.missing_data_folder, "missing_sas_formatted.csv")
        labeled_csv = os.path.join(self.missing_data_folder, "missing_sas_labeled.csv")
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file)
        df_csv = pd.read_csv(unformatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file, user_missing=True)
        df_csv = pd.read_csv(formatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        missing_user_values = {'var1':['A'],'var2': ['B'], 'var3':['C'], 'var4':['X'], 'var5':['Y'], 
        'var6':['Z'], 'var7':['_']}
        self.assertDictEqual(meta.missing_user_values, missing_user_values)

        df_sas, meta = pyreadstat.read_sas7bdat(sas_file,
                            catalog_file=cat_file, user_missing=True,
                            formats_as_category=False)
        df_csv = pd.read_csv(labeled_csv)
        self.assertTrue(df_sas.equals(df_csv))

def test_sas7bdat_metaonly(self):

        df, meta = pyreadstat.read_sas7bdat(os.path.join(self.basic_data_folder, "sample.sas7bdat"))
        df2, meta2 = pyreadstat.read_sas7bdat(os.path.join(self.basic_data_folder, "sample.sas7bdat"), metadataonly=True)
        self.assertTrue(df2.empty)
        self.assertTrue(meta.number_columns == meta2.number_columns)
        self.assertTrue(meta.number_rows == meta2.number_rows)
        self.assertTrue(meta.column_names == meta2.column_names)
        self.assertTrue(meta.column_labels == meta2.column_labels)

def test_sas_user_missing(self):
        sas_file = os.path.join(self.missing_data_folder, "missing_test.sas7bdat")
        cat_file = os.path.join(self.missing_data_folder, "missing_formats.sas7bcat")
        unformatted_csv = os.path.join(self.missing_data_folder, "missing_unformatted.csv")
        formatted_csv = os.path.join(self.missing_data_folder, "missing_sas_formatted.csv")
        labeled_csv = os.path.join(self.missing_data_folder, "missing_sas_labeled.csv")
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file)
        df_csv = pd.read_csv(unformatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file, user_missing=True)
        df_csv = pd.read_csv(formatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        missing_user_values = {'var1':['A'],'var2': ['B'], 'var3':['C'], 'var4':['X'], 'var5':['Y'], 
        'var6':['Z'], 'var7':['_']}
        self.assertDictEqual(meta.missing_user_values, missing_user_values)

        df_sas, meta = pyreadstat.read_sas7bdat(sas_file,
                            catalog_file=cat_file, user_missing=True,
                            formats_as_category=False)
        df_csv = pd.read_csv(labeled_csv)
        self.assertTrue(df_sas.equals(df_csv))

def test_sas7bdat_usecols(self):

        df, meta = pyreadstat.read_sas7bdat(os.path.join(self.basic_data_folder, "sample.sas7bdat"), usecols=self.usecols)
        self.assertTrue(df.equals(self.df_usecols))
        self.assertTrue(meta.number_columns == len(self.usecols))
        self.assertTrue(meta.column_names == self.usecols)

def test_sas_user_missing(self):
        sas_file = os.path.join(self.missing_data_folder, "missing_test.sas7bdat")
        cat_file = os.path.join(self.missing_data_folder, "missing_formats.sas7bcat")
        unformatted_csv = os.path.join(self.missing_data_folder, "missing_unformatted.csv")
        formatted_csv = os.path.join(self.missing_data_folder, "missing_sas_formatted.csv")
        labeled_csv = os.path.join(self.missing_data_folder, "missing_sas_labeled.csv")
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file)
        df_csv = pd.read_csv(unformatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file, user_missing=True)
        df_csv = pd.read_csv(formatted_csv)
        self.assertTrue(df_sas.equals(df_csv))
        missing_user_values = {'var1':['A'],'var2': ['B'], 'var3':['C'], 'var4':['X'], 'var5':['Y'], 
        'var6':['Z'], 'var7':['_']}
        self.assertDictEqual(meta.missing_user_values, missing_user_values)

        df_sas, meta = pyreadstat.read_sas7bdat(sas_file,
                            catalog_file=cat_file, user_missing=True,
                            formats_as_category=False)
        df_csv = pd.read_csv(labeled_csv)
        self.assertTrue(df_sas.equals(df_csv))

def test_sas7bdat_nodates(self):
        df, meta = pyreadstat.read_sas7bdat(os.path.join(self.basic_data_folder, "sample.sas7bdat"), disable_datetime_conversion=True)
        self.assertTrue(df.equals(self.df_nodates_sastata))

def test_sas_dates(self):

        sas_file = os.path.join(self.basic_data_folder, "dates.sas7bdat")
        df_sas, meta = pyreadstat.read_sas7bdat(sas_file)
        self.assertTrue(df_sas.equals(self.df_sas_dates))