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

refactoring

Browse Source
This commit is contained in:
frankknoll
2022-02-07 14:47:52 +01:00
parent 33dd6b229e
commit b097a2c74b
1 changed files with 27 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
HowBadIsMyBatch.ipynb
View File

@@ -172,7 +172,31 @@
"class AggregationHelper:\n",
"\n",
" @staticmethod\n",
" def aggregateAndFlattenColumnsAndRenameColumns(groupBy, aggFunctionsByColumn, columnNameMappingsDict):\n",
" def createSummationTable(\n",
" groupBy,\n",
" # FK-TODO: rename \"ADRs\" and \"Total reports\" to \"Total Number of Adverse Reaction Reports\" in all places\n",
" columnNameMappingsDict = {\n",
" \"DIED_size\": \"ADRs\",\n",
" \"DIED_sum\": \"DEATHS\",\n",
" \"L_THREAT_sum\": \"LIFE THREATENING ILLNESSES\",\n",
" \"DISABLE_sum\": \"DISABILITIES\",\n",
" 'HOSPITAL_sum': 'HOSPITALISATIONS',\n",
" 'ER_VISIT_sum': 'EMERGENCY ROOM OR DOCTOR VISITS'\n",
" }):\n",
" return AggregationHelper._aggregateAndFlattenColumnsAndRenameColumns(\n",
" groupBy = groupBy,\n",
" aggFunctionsByColumn = {\n",
" 'DIED': ['sum', 'size'],\n",
" 'L_THREAT': 'sum',\n",
" 'DISABLE': 'sum',\n",
" 'HOSPITAL': 'sum',\n",
" 'ER_VISIT': 'sum'\n",
" },\n",
" columnNameMappingsDict = columnNameMappingsDict)\n",
"\n",
" # FK-TODO: inline method\n",
" @staticmethod\n",
" def _aggregateAndFlattenColumnsAndRenameColumns(groupBy, aggFunctionsByColumn, columnNameMappingsDict):\n",
" aggregatedDataFrame = groupBy.agg(aggFunctionsByColumn)\n",
" AggregationHelper._flattenColumns(aggregatedDataFrame)\n",
" return aggregatedDataFrame.rename(columns = columnNameMappingsDict)\n",
@@ -198,34 +222,11 @@
"\n",
" # create table from https://www.howbadismybatch.com/combined.html\n",
" def createSevereEffectsBatchCodeTable(self):\n",
" batchCodeTable = BatchCodeTableHelper.createSummationTable(self.dataFrame.groupby('VAX_LOT'))\n",
" batchCodeTable = AggregationHelper.createSummationTable(self.dataFrame.groupby('VAX_LOT'))\n",
" batchCodeTable = batchCodeTable[['ADRs', 'DEATHS', 'DISABILITIES', 'LIFE THREATENING ILLNESSES', 'HOSPITALISATIONS', 'EMERGENCY ROOM OR DOCTOR VISITS']]\n",
" batchCodeTable = batchCodeTable.sort_values(by = 'ADRs', ascending = False)\n",
" return self._addCompanyColumn(batchCodeTable, self._createCompanyByBatchCodeTable())\n",
"\n",
" @staticmethod\n",
" def createSummationTable(\n",
" groupBy,\n",
" # FK-TODO: rename \"ADRs\" and \"Total reports\" to \"Total Number of Adverse Reaction Reports\" in all places\n",
" columnNameMappingsDict = {\n",
" \"DIED_size\": \"ADRs\",\n",
" \"DIED_sum\": \"DEATHS\",\n",
" \"L_THREAT_sum\": \"LIFE THREATENING ILLNESSES\",\n",
" \"DISABLE_sum\": \"DISABILITIES\",\n",
" 'HOSPITAL_sum': 'HOSPITALISATIONS',\n",
" 'ER_VISIT_sum': 'EMERGENCY ROOM OR DOCTOR VISITS'\n",
" }):\n",
" return AggregationHelper.aggregateAndFlattenColumnsAndRenameColumns(\n",
" groupBy = groupBy,\n",
" aggFunctionsByColumn = {\n",
" 'DIED': ['sum', 'size'],\n",
" 'L_THREAT': 'sum',\n",
" 'DISABLE': 'sum',\n",
" 'HOSPITAL': 'sum',\n",
" 'ER_VISIT': 'sum'\n",
" },\n",
" columnNameMappingsDict = columnNameMappingsDict)\n",
"\n",
" def _addCompanyColumn(self, batchCodeTable, companyByBatchCodeTable):\n",
" return pd.merge(\n",
" batchCodeTable,\n",
@@ -290,7 +291,7 @@
" # FK-TODO: DRY because it generates a subset of BatchCodeTableHelper.createSevereEffectsBatchCodeTable()\n",
" @staticmethod\n",
" def _getDoseTable(dataFrame):\n",
" doseTable = BatchCodeTableHelper.createSummationTable(\n",
" doseTable = AggregationHelper.createSummationTable(\n",
" dataFrame,\n",
" columnNameMappingsDict = {\n",
" \"DIED_size\": \"Total reports\",\n",