francois/HowBadIsMyBatch
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

refining VaersDescrReaderTest

Browse Source
This commit is contained in:
frankknoll
2023-01-18 17:26:25 +01:00
parent fc62639a1b
commit bc87b9d34d
2 changed files with 23 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
src/VaersDescrReader.py
View File

@@ -31,7 +31,11 @@ class VaersDescrReader:
return self._read_csv( return self._read_csv(
file = file, file = file,
usecols = ['VAERS_ID', 'VAX_DOSE_SERIES', 'VAX_TYPE', 'VAX_MANU', 'VAX_LOT'], usecols = ['VAERS_ID', 'VAX_DOSE_SERIES', 'VAX_TYPE', 'VAX_MANU', 'VAX_LOT'],
dtype = {"VAX_DOSE_SERIES": "string"}) dtype =
{
"VAX_DOSE_SERIES": "string",
"VAX_LOT": "string"
})
def _read_csv(self, file, **kwargs): def _read_csv(self, file, **kwargs):
return pd.read_csv( return pd.read_csv(

20
src/VaersDescrReaderTest.py
View File

@@ -15,11 +15,21 @@ class VaersDescrReaderTest(unittest.TestCase):
vaersDescr = vaersDescrReader.readVaersDescrForYear(2023) vaersDescr = vaersDescrReader.readVaersDescrForYear(2023)
# Then # Then
RECVDATE = pd.to_datetime('01/01/2023', format = "%m/%d/%Y") assert_frame_equal(
dataFrameExpected = TestHelper.createDataFrame( vaersDescr['VAERSDATA'],
TestHelper.createDataFrame(
columns = ['RECVDATE', 'DIED', 'L_THREAT', 'ER_VISIT', 'HOSPITAL', 'DISABLE', 'SPLTTYPE'], columns = ['RECVDATE', 'DIED', 'L_THREAT', 'ER_VISIT', 'HOSPITAL', 'DISABLE', 'SPLTTYPE'],
data = [ [RECVDATE, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]], data = [ [pd.to_datetime('01/01/2023', format = "%m/%d/%Y"), np.nan, np.nan, np.nan, np.nan, np.nan, np.nan]],
index = pd.Index( index = pd.Index(
name = 'VAERS_ID', name = 'VAERS_ID',
data=[2547730])) data=[2547730])),
assert_frame_equal(vaersDescr['VAERSDATA'], dataFrameExpected, check_dtype=False) check_dtype = False)
assert_frame_equal(
vaersDescr['VAERSVAX'],
TestHelper.createDataFrame(
columns = ['VAX_TYPE', 'VAX_MANU', 'VAX_LOT', 'VAX_DOSE_SERIES'],
data = [ ['COVID19', 'JANSSEN', '1808982', 'UNK']],
index = pd.Index(
name = 'VAERS_ID',
data=[2547730])),
check_dtype = False)