{ "Resources": { "covid19AE266CCE": { "Type": "AWS::Glue::Database", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseInput": { "Name": "covid-19" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid-19/Resource" } }, "jhuconsolidated": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Johns Hopkins University Consolidated data on COVID-19 cases, sourced from Enigma", "Name": "enigma_jhu", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "state and county two digits code", "Name": "fips", "Type": "string" }, { "Comment": "county name", "Name": "admin2", "Type": "string" }, { "Comment": "province name or state name", "Name": "province_state", "Type": "string" }, { "Comment": "country name or region name", "Name": "country_region", "Type": "string" }, { "Comment": "last update timestamp", "Name": "last_update", "Type": "string" }, { "Comment": "location (latitude)", "Name": "latitude", "Type": "double" }, { "Comment": "location (longitude)", "Name": "longitude", "Type": "double" }, { "Comment": "number of confirmed cases", "Name": "confirmed", "Type": "int" }, { "Comment": "number of deaths", "Name": "deaths", "Type": "int" }, { "Comment": "number of recovered patients", "Name": "recovered", "Type": "int" }, { "Comment": "number of active cases", "Name": "active", "Type": "int" }, { "Comment": "county name+state name+country name", "Name": "combined_key", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/archived/enigma-jhu/json", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "active,admin2,combined_key,confirmed,country_region,deaths,fips,last_update,latitude,longitude,province_state,recovered" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/jhu_consolidated" } }, "jhutimeseries": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Johns Hopkins University data on COVID-19 cases, sourced from Enigma", "Name": "enigma_jhu_timeseries", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "", "Name": "uid", "Type": "bigint" }, { "Comment": "state and county two digits code", "Name": "fips", "Type": "string" }, { "Comment": "", "Name": "iso2", "Type": "string" }, { "Comment": "", "Name": "iso3", "Type": "string" }, { "Comment": "", "Name": "code3", "Type": "int" }, { "Comment": "county name", "Name": "admin2", "Type": "string" }, { "Comment": "location (latitude)", "Name": "latitude", "Type": "double" }, { "Comment": "location (longitude)", "Name": "longitude", "Type": "double" }, { "Comment": "province name or state name", "Name": "province_state", "Type": "string" }, { "Comment": "country name or region name", "Name": "country_region", "Type": "string" }, { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "number of confirmed cases", "Name": "confirmed", "Type": "int" }, { "Comment": "number of deaths", "Name": "deaths", "Type": "int" }, { "Comment": "number of recovered patients", "Name": "recovered", "Type": "int" }, { "Comment": "number of active cases", "Name": "active", "Type": "int" }, { "Comment": "county name+state name+country name", "Name": "combined_key", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/archived/enigma-jhu-timeseries/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "admin2,code3,confirmed,country_region,date,deaths,fips,iso2,iso3,latitude,longitude,province_state,recovered,uid" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/jhu_timeseries" } }, "tableaujhu": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Johns Hopkins University data on COVID-19 cases, sourced from Tableau", "Name": "tableau_jhu", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "Confirmed, Deaths, etc.", "Name": "case_type", "Type": "string" }, { "Comment": "number of cases", "Name": "cases", "Type": "bigint" }, { "Comment": "difference in cases vs the previous day", "Name": "difference", "Type": "bigint" }, { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "country name or region name", "Name": "country_region", "Type": "string" }, { "Comment": "province name or state name", "Name": "province_state", "Type": "string" }, { "Comment": "county name", "Name": "admin2", "Type": "string" }, { "Comment": "county name+state name+country name", "Name": "combined_key", "Type": "string" }, { "Comment": "state and county two digits code", "Name": "fips", "Type": "string" }, { "Comment": "location (latitude)", "Name": "lat", "Type": "double" }, { "Comment": "location (longitude)", "Name": "long", "Type": "double" }, { "Comment": "source table name", "Name": "table_names", "Type": "string" }, { "Comment": "timestamp when the data was generated for this data load", "Name": "prep_flow_runtime", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/archived/tableau-jhu/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "Admin2,Case_Type,Cases,Combined_Key,Country_Region,Date,Difference,FIPS,Lat,Long,Prep_Flow_Runtime,Province_State,Table_Names" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/tableau_jhu" } }, "alleninstitutemetadata": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Metadata on papers pulled from the Allen Institute. The 'sha' column indicates the paper id", "Name": "alleninstitute_metadata", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "unique identifier coming from CORD (COVID-19 Open Research Dataset)", "Name": "cord_uid", "Type": "string" }, { "Comment": "hash for the paper id. can now include multiple files (some PMC files have multiple associated PDFs)", "Name": "sha", "Type": "string" }, { "Comment": "the data source, currently: PMC, biorxiv, CZI, or medrxiv", "Name": "source_x", "Type": "string" }, { "Comment": "the title of the paper", "Name": "title", "Type": "string" }, { "Comment": "doi id for the paper", "Name": "doi", "Type": "string" }, { "Comment": "pmc id for the paper", "Name": "pmcid", "Type": "string" }, { "Comment": "pubmed id for the paper", "Name": "pubmed_id", "Type": "string" }, { "Comment": "license associated to the paper", "Name": "license", "Type": "string" }, { "Comment": "abstract of the paper", "Name": "abstract", "Type": "string" }, { "Comment": "When the paper was published. Some papers are at the day level, while others are at the year.", "Name": "publish_time", "Type": "string" }, { "Name": "authors", "Type": "string" }, { "Comment": "journal in which the paper was published", "Name": "journal", "Type": "string" }, { "Comment": "paper id in microsoft academic (if applicable)", "Name": "microsoft academic paper id", "Type": "string" }, { "Comment": "covidence number from WHO", "Name": "who #covidence", "Type": "string" }, { "Comment": "whether the full text of the paper is available", "Name": "has_full_text", "Type": "boolean" }, { "Comment": "which S3 folder/prefix the full text is in", "Name": "full_text_file", "Type": "string" }, { "Comment": "url of the journal or paper", "Name": "url", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/alleninstitute/CORD19/json/metadata", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "Microsoft Academic Paper ID,WHO #Covidence,abstract,authors,cord_uid,doi,full_text_file,has_full_text,journal,license,pmcid,publish_time,pubmed_id,sha,source_x,title,url" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/alleninstitute_metadata" } }, "alleninstitutecomprehendmedical": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Comprehend Medical results run against Allen Institute data on medical papers.", "Name": "alleninstitute_comprehend_medical", "Parameters": { "compressionType": "none", "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "the id of the paper", "Name": "paper_id", "Type": "string" }, { "Comment": "", "Name": "date", "Type": "array" }, { "Comment": "All medical conditions listed. Includes present illness, reason for visit, and medical history", "Name": "dx_name", "Type": "array" }, { "Comment": "Procedures performed on a patient for diagnostic, measurement, screening, or rating that might have a resulting value", "Name": "test_name", "Type": "array" }, { "Comment": "Interventions as a one-time action performed on the patient to treat a medical condition or to provide patient care", "Name": "procedure_name", "Type": "array" }, { "Comment": "Any phone, fax, or pager number. Excludes named phone numbers, such as 1-800-QUIT-NOW and 911", "Name": "phone_or_fax", "Type": "array" }, { "Comment": "The date a test was performed", "Name": "time_to_test_name", "Type": "array" }, { "Comment": "", "Name": "url", "Type": "array" }, { "Comment": "Non-brand name, ingredient name, or formula mixture of the medication or therapeutic agent", "Name": "generic_name", "Type": "array" }, { "Comment": "All names. Typically, names of the patient, family, or provider", "Name": "name", "Type": "array" }, { "Comment": "The copyrighted brand name of the medication or therapeutic agent", "Name": "brand_name", "Type": "array" }, { "Comment": "All geographical subdivisions of an address of any facility, named medical facilities, or wards within a facility", "Name": "address", "Type": "array" }, { "Comment": "Social security number, medical record number, facility identification number, clinical trial number, certificate or license number, vehicle or device number. This includes any biometric number of the patient, place of care, or provider.", "Name": "id", "Type": "array" }, { "Comment": "Interventions performed over a span of time for combating a disease or disorder. This includes groupings of medications, such as antivirals and vaccinations", "Name": "treatment_name", "Type": "array" }, { "Comment": "Body systems, anatomic locations or regions, and body sites", "Name": "system_organ_site", "Type": "array" }, { "Comment": "The date a treatment was administered", "Name": "time_to_treatment_name", "Type": "array" }, { "Comment": "The date a medical condition occurred", "Name": "time_to_dx_name", "Type": "array" }, { "Comment": "The date a medication was taken", "Name": "time_to_medication_name", "Type": "array" }, { "Comment": "The date a procedure was performed", "Name": "time_to_procedure_name", "Type": "array" }, { "Comment": "Any profession or employer that pertains to the patient or the patient's family. It does include the profession of the clinician mentioned in the note", "Name": "profession", "Type": "array" }, { "Comment": "Any email address", "Name": "email", "Type": "array" }, { "Comment": "All components of age, spans of age, or any age mentioned. This includes those of a patient, family members, or others. The default is in years unless otherwise noted", "Name": "age", "Type": "array" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/alleninstitute/CORD19/comprehendmedical/", "NumberOfBuckets": -1, "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "Parameters": { "compressionType": "none", "classification": "json", "typeOfData": "file" }, "SerdeInfo": { "Parameters": { "paths": "ADDRESS,AGE,BRAND_NAME,DATE,DX_NAME,EMAIL,GENERIC_NAME,ID,NAME,PHONE_OR_FAX,PROCEDURE_NAME,PROFESSION,SYSTEM_ORGAN_SITE,TEST_NAME,TIME_TO_DX_NAME,TIME_TO_MEDICATION_NAME,TIME_TO_PROCEDURE_NAME,TIME_TO_TEST_NAME,TIME_TO_TREATMENT_NAME,TREATMENT_NAME,URL,paper_id" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/alleninstitute_comprehend_medical" } }, "covidtestingusdaily": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "USA total test daily trend. Sourced from covidtracking.com via REARC", "Name": "covid_testing_us_daily", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "number of states and territories included", "Name": "states", "Type": "bigint" }, { "Comment": "number of positive cases", "Name": "positive", "Type": "double" }, { "Comment": "number of negative cases", "Name": "negative", "Type": "double" }, { "Comment": "positive + negative cases", "Name": "posNeg", "Type": "double" }, { "Comment": "tests pending results", "Name": "pending", "Type": "double" }, { "Comment": "number of hospitalized patients", "Name": "hospitalized", "Type": "double" }, { "Comment": "number of deaths", "Name": "death", "Type": "double" }, { "Comment": "total tests", "Name": "total", "Type": "double" }, { "Comment": "hash value", "Name": "hash", "Type": "string" }, { "Comment": "last data sync", "Name": "dateChecked", "Type": "string" }, { "Comment": "total test results", "Name": "totalTestResults", "Type": "double" }, { "Comment": "increase in deaths vs previous day", "Name": "deathIncrease", "Type": "double" }, { "Comment": "increase in hospitalized patients vs previous day", "Name": "hospitalizedIncrease", "Type": "double" }, { "Comment": "increase in negative cases vs previous day", "Name": "negativeIncrease", "Type": "double" }, { "Comment": "increase in positive cases vs previous day", "Name": "positiveIncrease", "Type": "double" }, { "Comment": "increase in total results vs previous day", "Name": "totalTestResultsIncrease", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-testing-data/json/us_daily", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,dateChecked,death,deathIncrease,hash,hospitalized,hospitalizedIncrease,negative,negativeIncrease,pending,posNeg,positive,positiveIncrease,states,total,totalTestResults,totalTestResultsIncrease" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_testing_us_daily" } }, "covidtestingstatesdaily": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "USA total test daily trend by state. Sourced from covidtracking.com via REARC", "Name": "covid_testing_states_daily", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file", "areColumnsQuoted": "false", "columnsOrdered": "true", "delimiter": ",", "skip.header.line.count": "1" }, "StorageDescriptor": { "Columns": [ { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "US State", "Name": "state", "Type": "string" }, { "Comment": "number of positive cases", "Name": "positive", "Type": "double" }, { "Comment": "number of negative cases", "Name": "negative", "Type": "double" }, { "Comment": "tests pending results", "Name": "pending", "Type": "double" }, { "Comment": "number of hospitalized patients", "Name": "hospitalized", "Type": "double" }, { "Comment": "number of deaths", "Name": "death", "Type": "double" }, { "Comment": "total tests", "Name": "total", "Type": "double" }, { "Comment": "hash value", "Name": "hash", "Type": "string" }, { "Comment": "last data sync", "Name": "dateChecked", "Type": "string" }, { "Comment": "total test results", "Name": "totalTestResults", "Type": "double" }, { "Comment": "fips code", "Name": "fips", "Type": "string" }, { "Comment": "increase in deaths vs previous day", "Name": "deathIncrease", "Type": "double" }, { "Comment": "increase in hospitalized patients vs previous day", "Name": "hospitalizedIncrease", "Type": "double" }, { "Comment": "increase in negative cases vs previous day", "Name": "negativeIncrease", "Type": "double" }, { "Comment": "increase in positive cases vs previous day", "Name": "positiveIncrease", "Type": "double" }, { "Comment": "increase in total results vs previous day", "Name": "totalTestResultsIncrease", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-testing-data/json/states_daily", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,dateChecked,death,deathIncrease,fips,hash,hospitalized,hospitalizedIncrease,negative,negativeIncrease,pending,positive,positiveIncrease,state,total,totalTestResults,totalTestResultsIncrease" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_testing_states_daily" } }, "covidtestingustotal": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "USA total tests. Sourced from covidtracking.com via REARC", "Name": "covid_testing_us_total", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "number of positive cases", "Name": "positive", "Type": "double" }, { "Comment": "number of negative cases", "Name": "negative", "Type": "double" }, { "Comment": "positive + negative cases", "Name": "posNeg", "Type": "double" }, { "Comment": "number of hospitalized patients", "Name": "hospitalized", "Type": "double" }, { "Comment": "number of deaths", "Name": "death", "Type": "double" }, { "Comment": "total tests", "Name": "total", "Type": "double" }, { "Comment": "hash value", "Name": "hash", "Type": "string" }, { "Comment": "eporting date", "Name": "lastModified", "Type": "string" }, { "Comment": "total test results", "Name": "totalTestResults", "Type": "double" }, { "Comment": "", "Name": "notes", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-testing-data/json/us-total-latest", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "death,hash,hospitalized,lastModified,negative,notes,posNeg,positive,total,totalTestResults" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_testing_us_total" } }, "stateabbrevstable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Lookup table for US state abbreviations", "Name": "us_state_abbreviations", "Parameters": { "has_encrypted_data": false, "classification": "csv", "areColumnsQuoted": "false", "typeOfData": "file", "columnsOrdered": "true", "delimiter": ",", "skip.header.line.count": "1" }, "StorageDescriptor": { "Columns": [ { "Comment": "State name", "Name": "state", "Type": "string" }, { "Comment": "abbreviation", "Name": "abbreviation", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/static-datasets/csv/state-abv", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "field.delim": "," }, "SerializationLibrary": "org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/state_abbrevs_table" } }, "countypoptable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Lookup table for population for each county based on recent census data", "Name": "county_populations", "Parameters": { "has_encrypted_data": false, "classification": "csv", "areColumnsQuoted": "false", "typeOfData": "file", "columnsOrdered": "true", "delimiter": ",", "skip.header.line.count": "1" }, "StorageDescriptor": { "Columns": [ { "Comment": "geo id", "Name": "Id", "Type": "string" }, { "Comment": "geo id2", "Name": "Id2", "Type": "string" }, { "Comment": "county name", "Name": "county", "Type": "string" }, { "Comment": "state name", "Name": "state", "Type": "string" }, { "Comment": "", "Name": "Population Estimate 2018", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/static-datasets/csv/CountyPopulation", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "field.delim": "," }, "SerializationLibrary": "org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/county_pop_table" } }, "countrycodetable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Lookup table for country codes", "Name": "country_codes", "Parameters": { "has_encrypted_data": false, "classification": "csv", "areColumnsQuoted": "false", "typeOfData": "file", "columnsOrdered": "true", "delimiter": ",", "skip.header.line.count": "1" }, "StorageDescriptor": { "Columns": [ { "Comment": "geo id", "Name": "Country", "Type": "string" }, { "Comment": "geo id2", "Name": "Alpha-2 code", "Type": "string" }, { "Comment": "state name", "Name": "Alpha-3 code", "Type": "string" }, { "Comment": "", "Name": "Numeric code", "Type": "bigint" }, { "Comment": "location (latitude)", "Name": "Latitude", "Type": "bigint" }, { "Comment": "location (longitude)", "Name": "Longitude", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/static-datasets/csv/countrycode", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "field.delim": "," }, "SerializationLibrary": "org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/country_code_table" } }, "nytimescountiestable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Data on COVID-19 cases from NY Times at US county level", "Name": "nytimes_counties", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "", "Name": "county", "Type": "string" }, { "Comment": "", "Name": "state", "Type": "string" }, { "Comment": "FIPS code", "Name": "fips", "Type": "string" }, { "Comment": "# confirmed cases", "Name": "cases", "Type": "bigint" }, { "Comment": "# deaths", "Name": "deaths", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-nyt-data-in-usa/json/us-counties", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,county,state,fips,cases,deaths" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/nytimes_counties_table" } }, "nytimesstatestable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Data on COVID-19 cases from NY Times at US state level", "Name": "nytimes_states", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "", "Name": "state", "Type": "string" }, { "Comment": "FIPS code", "Name": "fips", "Type": "string" }, { "Comment": "# confirmed cases", "Name": "cases", "Type": "bigint" }, { "Comment": "# deaths", "Name": "deaths", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-nyt-data-in-usa/json/us-states", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,state,fips,cases,deaths" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/nytimes_states_table" } }, "hospitalbedstable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Data on hospital beds and their utilization in the US. Sourced from Definitive Healthcare via rearc", "Name": "hospital_beds", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "unique id for the record", "Name": "objectid", "Type": "int" }, { "Name": "hospital_name", "Type": "string" }, { "Comment": "Short Term Acute Care Hospital (STAC), Critical Access Hospital (CAH), Long Term Acute Care Hospitals, Children’s Hospitals, Veteran's Affairs (VA) Hospital or Department of Defense (DoD) Hospital", "Name": "hospital_type", "Type": "string" }, { "Name": "hq_address", "Type": "string" }, { "Name": "hq_address1", "Type": "string" }, { "Name": "hq_city", "Type": "string" }, { "Name": "hq_state", "Type": "string" }, { "Name": "hq_zip_code", "Type": "string" }, { "Name": "county_name", "Type": "string" }, { "Name": "state_name", "Type": "string" }, { "Name": "state_fips", "Type": "string" }, { "Name": "cnty_fips", "Type": "string" }, { "Name": "fips", "Type": "string" }, { "Comment": "maximum number of beds for which a hospital holds a license to operate", "Name": "num_licensed_beds", "Type": "int" }, { "Comment": "adult bed, pediatric bed, birthing room, or newborn ICU bed (excluding newborn bassinets) maintained in a patient care area for lodging patients in acute, long term, or domiciliary areas of the hospital.", "Name": "num_staffed_beds", "Type": "int" }, { "Comment": "ICU beds, burn ICU beds, surgical ICU beds, premature ICU beds, neonatal ICU beds, pediatric ICU beds, psychiatric ICU beds, trauma ICU beds, and Detox ICU beds", "Name": "num_icu_beds", "Type": "int" }, { "Comment": "calculated based on metrics from the Medicare Cost Report: Bed Utilization Rate = Total Patient Days (excluding nursery days)/Bed Days Available", "Name": "bed_utilization", "Type": "double" }, { "Comment": "computed by subtracting “Number of Staffed Beds from Number of Licensed beds” (Licensed Beds – Staffed Beds). This would provide insights into scenario planning for when staff can be shifted around to increase available bed capacity as needed.", "Name": "potential_increase_in_bed_capac", "Type": "int" }, { "Comment": "hospital location (latitude)", "Name": "latitude", "Type": "double" }, { "Comment": "hospital location (longitude)", "Name": "longtitude", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-usa-hospital-beds", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "BED_UTILIZATION,CNTY_FIPS,COUNTY_NAME,FIPS,HOSPITAL_NAME,HOSPITAL_TYPE,HQ_ADDRESS,HQ_ADDRESS1,HQ_CITY,HQ_STATE,HQ_ZIP_CODE,NUM_ICU_BEDS,NUM_LICENSED_BEDS,NUM_STAFFED_BEDS,OBJECTID,Potential_Increase_In_Bed_Capac,STATE_FIPS,STATE_NAME,latitude,longtitude" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/hospital_beds_table" } }, "covidcastdatatable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "CMU Delphi's COVID-19 Surveillance Data", "Name": "covidcast_data", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "name of upstream data source (e.g., fb-survey, google-survey, ght, quidel, doctor-visits)", "Name": "data_source", "Type": "string" }, { "Comment": "name of signal derived from upstream data", "Name": "signal", "Type": "string" }, { "Comment": "spatial resolution of the signal (e.g., county, hrr, msa, dma, state)", "Name": "geo_type", "Type": "string" }, { "Comment": "time unit (e.g., date) over which underlying events happened", "Name": "time_value", "Type": "int" }, { "Comment": "unique code for each location, depending on geo_type (county -> FIPS 6-4 code, HRR -> HRR number, MSA -> CBSA code, DMA -> DMA code, state -> two-letter state code), or * for all", "Name": "geo_value", "Type": "string" }, { "Comment": "trend classifier (+1 -> increasing, 0 steady or not determined, -1 -> decreasing)", "Name": "direction", "Type": "int" }, { "Comment": "value (statistic) derived from the underlying data source", "Name": "value", "Type": "double" }, { "Comment": "standard error of the statistic with respect to its sampling distribution", "Name": "stderr", "Type": "double" }, { "Comment": "number of data points used in computing the statistic", "Name": "sample_size", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covidcast/json/data/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "data_source,direction,geo_type,geo_value,sample_size,signal,stderr,time_value,value" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covidcast_data_table" } }, "covidcastmetatable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "CMU Delphi's COVID-19 Surveillance Metadata", "Name": "covidcast_metadata", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "name of upstream data source (e.g., fb-survey, google-survey, ght, quidel, doctor-visits)", "Name": "data_source", "Type": "string" }, { "Comment": "name of signal derived from upstream data", "Name": "signal", "Type": "string" }, { "Comment": "temporal resolution of the signal (e.g., day, week)", "Name": "time_type", "Type": "string" }, { "Comment": "geographic resolution (e.g. county, hrr, msa, dma, state)", "Name": "geo_type", "Type": "string" }, { "Comment": "minimum time (e.g., 20200406)", "Name": "min_time", "Type": "int" }, { "Comment": "maximum time (e.g., 20200413)", "Name": "max_time", "Type": "int" }, { "Name": "num_locations", "Type": "int" }, { "Name": "min_value", "Type": "double" }, { "Name": "max_value", "Type": "double" }, { "Name": "mean_value", "Type": "double" }, { "Name": "stdev_value", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covidcast/json/metadata/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "data_source,geo_type,max_time,max_value,mean_value,min_time,min_value,num_locations,signal,stdev_value,time_type" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covidcast_meta_table" } }, "covidknowledgegraphedgestable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_edges", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "from", "Type": "string" }, { "Name": "to", "Type": "string" }, { "Name": "score", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/edges/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "from,id,label,score,to" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_edges_table" } }, "covidknowledgegraphnodesconcepttable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_nodes_concept", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "entity", "Type": "string" }, { "Name": "concept", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/nodes/concept_nodes/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "concept,entity,id,label" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_nodes_concept_table" } }, "covidknowledgegraphnodesinstitutiontable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_nodes_institution", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "institution", "Type": "string" }, { "Name": "country", "Type": "string" }, { "Name": "settlement", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/nodes/institution_nodes/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "country,id,institution,label,settlement" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_nodes_institution_table" } }, "covidknowledgegraphnodespaperauthortable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_nodes_author", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "first", "Type": "string" }, { "Name": "last", "Type": "string" }, { "Name": "full_name", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/nodes/paper_author_nodes/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "first,full_name,id,label,last" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_nodes_paper_author_table" } }, "covidknowledgegraphnodespapertable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_nodes_paper", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "doi", "Type": "string" }, { "Name": "sha_code", "Type": "string" }, { "Name": "publish_time", "Type": "string" }, { "Name": "source", "Type": "string" }, { "Name": "title", "Type": "string" }, { "Name": "year", "Type": "int" }, { "Name": "pmcid", "Type": "string" }, { "Name": "reference", "Type": "boolean" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/nodes/paper_nodes/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "DOI,PMCID,SHA_code,id,label,publish_time,reference,source,title,year" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_nodes_paper_table" } }, "covidknowledgegraphnodestopictable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "AWS Knowledge Graph for COVID-19 data", "Name": "covid_knowledge_graph_nodes_topic", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "id", "Type": "string" }, { "Name": "label", "Type": "string" }, { "Name": "topic", "Type": "string" }, { "Name": "topic_num", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/covid_knowledge_graph/json/nodes/topic_nodes/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "id,label,topic,topic_num" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/covid_knowledge_graph_nodes_topic_table" } }, "predictionmodelsseverityindextable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Severity Index models. Sourced from Yu Group at UC Berkeley via Rearc.", "Name": "prediction_models_severity_index", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "severity_1-day", "Type": "int" }, { "Name": "severity_2-day", "Type": "int" }, { "Name": "severity_3-day", "Type": "int" }, { "Name": "severity_4-day", "Type": "int" }, { "Name": "severity_5-day", "Type": "int" }, { "Name": "severity_6-day", "Type": "int" }, { "Name": "severity_7-day", "Type": "int" }, { "Name": "total_deaths_hospital", "Type": "double" }, { "Name": "hospital_name", "Type": "string" }, { "Name": "cms_certification_number", "Type": "int" }, { "Name": "countyfips", "Type": "string" }, { "Name": "countyname", "Type": "string" }, { "Name": "statename", "Type": "string" }, { "Name": "system_affiliation", "Type": "string" }, { "Name": "latitude", "Type": "double" }, { "Name": "longitude", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-prediction-models/json/severity-index/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "cms_certification_number,countyfips,countyname,hospital_name,latitude,longitude,severity_1-day,severity_2-day,severity_3-day,severity_4-day,severity_5-day,severity_6-day,severity_7-day,statename,system_affiliation,total_deaths_hospital" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/prediction_models_severity_index_table" } }, "predictionmodelscountypredictionstable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "County-level Predictions Data. Sourced from Yu Group at UC Berkeley via Rearc.", "Name": "prediction_models_county_predictions", "Parameters": { "compressionType": "none", "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "countyfips", "Type": "string" }, { "Name": "countyname", "Type": "string" }, { "Name": "statename", "Type": "string" }, { "Name": "severity_county_5-day", "Type": "int" }, { "Name": "predicted_date", "Type": "string" }, { "Name": "predicted_deaths", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-prediction-models/json/county-predictions/", "NumberOfBuckets": -1, "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "Parameters": { "compressionType": "none", "classification": "json", "typeOfData": "file" }, "SerdeInfo": { "Parameters": {}, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/prediction_models_county_predictions_table" } }, "tableaudatahubtable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "COVID-19 data that has been gathered and unified from trusted sources to include the New York Times and the European CDC. Sourced from Tableau", "Name": "tableau_covid_datahub", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "country_short_name", "Type": "string" }, { "Name": "country_alpha_3_code", "Type": "string" }, { "Name": "country_alpha_2_code", "Type": "string" }, { "Name": "continent_name", "Type": "string" }, { "Name": "province_state_name", "Type": "string" }, { "Name": "county_name", "Type": "string" }, { "Name": "county_fips_number", "Type": "string" }, { "Comment": " Total (cumulative) positive cases", "Name": "people_positive_cases_count", "Type": "bigint" }, { "Name": "report_date", "Type": "string" }, { "Comment": "NYT for New York Times and EUDC for European Centre for Disease Prevention and Control", "Name": "data_source_name", "Type": "string" }, { "Comment": "total deaths for each day", "Name": "people_death_new_count", "Type": "bigint" }, { "Comment": "Total new positive cases for each day", "Name": "people_positive_new_cases_count", "Type": "bigint" }, { "Comment": "Total (cumulative) deaths", "Name": "people_death_count", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/tableau-covid-datahub/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "CONTINENT_NAME,COUNTRY_ALPHA_2_CODE,COUNTRY_ALPHA_3_CODE,COUNTRY_SHORT_NAME,COUNTY_FIPS_NUMBER,COUNTY_NAME,DATA_SOURCE_NAME,PEOPLE_DEATH_COUNT,PEOPLE_DEATH_NEW_COUNT,PEOPLE_POSITIVE_CASES_COUNT,PEOPLE_POSITIVE_NEW_CASES_COUNT,PROVINCE_STATE_NAME,REPORT_DATE" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/tableau_datahub_table" } }, "worldcasesdeathstestingtable": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Data on confirmed cases, deaths, and testing. Sourced from rearc.", "Name": "world_cases_deaths_testing", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "iso_code", "Type": "string" }, { "Name": "location", "Type": "string" }, { "Name": "date", "Type": "string" }, { "Name": "total_cases", "Type": "double" }, { "Name": "new_cases", "Type": "double" }, { "Name": "total_deaths", "Type": "double" }, { "Name": "new_deaths", "Type": "double" }, { "Name": "total_cases_per_million", "Type": "double" }, { "Name": "new_cases_per_million", "Type": "double" }, { "Name": "total_deaths_per_million", "Type": "double" }, { "Name": "new_deaths_per_million", "Type": "double" }, { "Name": "total_tests", "Type": "double" }, { "Name": "new_tests", "Type": "double" }, { "Name": "total_tests_per_thousand", "Type": "double" }, { "Name": "new_tests_per_thousand", "Type": "double" }, { "Name": "tests_units", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/rearc-covid-19-world-cases-deaths-testing/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,iso_code,location,new_cases,new_cases_per_million,new_deaths,new_deaths_per_million,new_tests,new_tests_per_thousand,tests_units,total_cases,total_cases_per_million,total_deaths,total_deaths_per_million,total_tests,total_tests_per_thousand" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/world_cases_deaths_testing_table" } }, "aspirevccrowdtracing": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Contact Tracing data from AspireVC", "Name": "aspirevc_crowd_tracing", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "Date that the scan occurred", "Name": "scandate", "Type": "string" }, { "Comment": "Unique Tokenized identifier of kiosk device", "Name": "scannerdeviceid", "Type": "string" }, { "Comment": "ZIP Code of Device Location", "Name": "scannerdevice_zipcode", "Type": "string" }, { "Comment": "Total number of people they can hold", "Name": "location_maxcapacity", "Type": "bigint" }, { "Comment": "Government enforced limit %", "Name": "location_capacity_enforcedlimit", "Type": "bigint" }, { "Comment": "Operating hours of device location", "Name": "location_op_hours", "Type": "string" }, { "Comment": "Unique Tokenized identifier of smart phone", "Name": "userdeviceid", "Type": "string" }, { "Comment": "Checkin or Checkout", "Name": "type_of_scan", "Type": "string" }, { "Comment": "Unique Tokenized identifier of user", "Name": "userid", "Type": "string" }, { "Comment": "Time in seconds for how long it took the user to complete the survey", "Name": "duration", "Type": "bigint" }, { "Comment": "Classification determined by the Go Pass system", "Name": "risklevel", "Type": "string" }, { "Comment": "The type of door being entered. 'Public' or 'Private'", "Name": "accesslevel", "Type": "string" }, { "Comment": "Device type used. 'Gate' or 'Personal' (Gate = Kiosk; Personal = Smart Phone) ", "Name": "readingtype", "Type": "string" }, { "Comment": "Temperature of user in F", "Name": "temp", "Type": "double" }, { "Comment": "O2 level of user", "Name": "o2", "Type": "bigint" }, { "Comment": "Past 2 weeks, have you had any of the following symptoms: shortness of breath, fever, loss of taste or smell, new cough?", "Name": "symptoms", "Type": "boolean" }, { "Comment": "Past 2 weeks, have you been diagnosed with COVID or are waiting for COVID test results?", "Name": "diagnosed", "Type": "boolean" }, { "Comment": "Past 2 weeks, have you been in contact with anyone who has been diagnosed with COVID or is waiting for COVID test results?", "Name": "contact", "Type": "boolean" }, { "Comment": "Past 2 weeks, have you been near anyone with the following symptoms: shortness of breath, fever, loss of taste or smell, new cough?", "Name": "near", "Type": "boolean" }, { "Comment": "User's mask usage score", "Name": "m_score", "Type": "double" }, { "Comment": "User's social distancing score", "Name": "s_score", "Type": "double" }, { "Comment": "Result of the scan. 'Approved' or 'Declined'", "Name": "result", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/aspirevc_crowd_tracing/json/data", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "ACCESSLEVEL,CONTACT,DIAGNOSED,DURATION,LOCATION_CAPACITY_ENFORCEDLIMIT,LOCATION_MAXCAPACITY,LOCATION_OP_HOURS,M_SCORE,NEAR,O2,READTYPE,RESULT,RISKLEVEL,SCANDATE,SCANNERDEVICEID,SCANNERDEVICE_ZIPCODE,SYMPTOMS,S_SCORE,TEMP,TYPE_OF_SCAN,USERDEVICEID,USERID" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/aspirevc_crowd_tracing" } }, "aspirevccrowdtracingzipcode3digits": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "3 digit zip code lookup for AspireVC tracing data", "Name": "aspirevc_crowd_tracing_zipcode_3digits", "Parameters": { "has_encrypted_data": false, "typeOfData": "file", "skip.header.line.count": "1", "classification": "csv" }, "StorageDescriptor": { "Columns": [ { "Comment": "3 digit zip", "Name": "zip", "Type": "string" }, { "Comment": "US Sate", "Name": "state", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/aspirevc_crowd_tracing/csv/zip-data/3digits/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "separatorChar": "," }, "SerializationLibrary": "org.apache.hadoop.hive.serde2.OpenCSVSerde" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/aspirevc_crowd_tracing_zipcode_3digits" } }, "enigmaaggregationglobal": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Aggregation of COVID-19 data from Our World in Data, The New York Times and The COVID Tracking project. All geographies combined. Sourced from Enigma", "Name": "enigma_aggregation_global", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "geographic_level", "Type": "string" }, { "Name": "country_name", "Type": "string" }, { "Name": "country_iso2", "Type": "string" }, { "Name": "country_iso3", "Type": "string" }, { "Name": "state_fips", "Type": "string" }, { "Name": "state_name", "Type": "string" }, { "Name": "county_fips", "Type": "string" }, { "Name": "county_name", "Type": "string" }, { "Name": "area_name", "Type": "string" }, { "Name": "lat", "Type": "double" }, { "Name": "long", "Type": "double" }, { "Name": "population", "Type": "double" }, { "Name": "date", "Type": "string" }, { "Name": "cases", "Type": "double" }, { "Name": "deaths", "Type": "double" }, { "Name": "tests", "Type": "double" }, { "Name": "tests_pending", "Type": "double" }, { "Name": "tests_negative", "Type": "double" }, { "Name": "tests_positive", "Type": "double" }, { "Name": "tests_units", "Type": "string" }, { "Name": "patients_icu", "Type": "double" }, { "Name": "patients_hosp", "Type": "double" }, { "Name": "patients_vent", "Type": "double" }, { "Name": "recovered", "Type": "double" }, { "Name": "version_timestamp", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/enigma-aggregation/json/global/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "area_name,cases,country_iso2,country_iso3,country_name,county_fips,county_name,date,deaths,geographic_level,lat,long,patients_hosp,patients_icu,patients_vent,population,recovered,state_fips,state_name,tests,tests_negative,tests_pending,tests_positive,tests_units,version_timestamp" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/enigma_aggregation_global" } }, "enigmaaggregationglobalcountries": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Aggregation of COVID-19 data from Our World in Data, The New York Times and The COVID Tracking project. Country level only. Sourced from Enigma", "Name": "enigma_aggregation_global_countries", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "country_name", "Type": "string" }, { "Name": "country_iso2", "Type": "string" }, { "Name": "country_iso3", "Type": "string" }, { "Name": "lat", "Type": "double" }, { "Name": "long", "Type": "double" }, { "Name": "population", "Type": "double" }, { "Name": "date", "Type": "string" }, { "Name": "cases", "Type": "double" }, { "Name": "deaths", "Type": "double" }, { "Name": "tests", "Type": "double" }, { "Name": "tests_units", "Type": "string" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/enigma-aggregation/json/global_countries/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "cases,country_iso2,country_iso3,country_name,date,deaths,lat,long,population,tests,tests_units" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/enigma_aggregation_global_countries" } }, "enigmaaggregationuscounties": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Aggregation of COVID-19 data from Our World in Data, The New York Times and The COVID Tracking project. US counties only. Sourced from Enigma", "Name": "enigma_aggregation_us_counties", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "state_fips", "Type": "string" }, { "Name": "state_name", "Type": "string" }, { "Name": "county_fips", "Type": "string" }, { "Name": "county_name", "Type": "string" }, { "Name": "area_name", "Type": "string" }, { "Name": "lat", "Type": "double" }, { "Name": "long", "Type": "double" }, { "Name": "date", "Type": "string" }, { "Name": "cases", "Type": "double" }, { "Name": "deaths", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/enigma-aggregation/json/us_counties/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "area_name,cases,county_fips,county_name,date,deaths,lat,long,state_fips,state_name" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/enigma_aggregation_us_counties" } }, "enigmaaggregationusstates": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Aggregation of COVID-19 data from Our World in Data, The New York Times and The COVID Tracking project. US states only. Sourced from Enigma", "Name": "enigma_aggregation_us_states", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Name": "state_fips", "Type": "string" }, { "Name": "state_name", "Type": "string" }, { "Name": "lat", "Type": "double" }, { "Name": "long", "Type": "double" }, { "Name": "date", "Type": "string" }, { "Name": "cases", "Type": "double" }, { "Name": "deaths", "Type": "double" }, { "Name": "tests", "Type": "double" }, { "Name": "tests_pending", "Type": "double" }, { "Name": "tests_negative", "Type": "double" }, { "Name": "tests_positive", "Type": "double" }, { "Name": "patients_icu", "Type": "double" }, { "Name": "patients_hosp", "Type": "double" }, { "Name": "patients_vent", "Type": "double" }, { "Name": "recovered", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/enigma-aggregation/json/us_states/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "cases,date,deaths,lat,long,patients_hosp,patients_icu,patients_vent,recovered,state_fips,state_name,tests,tests_negative,tests_pending,tests_positive" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/enigma_aggregation_us_states" } }, "ukcovid": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "COVID-19 case and testing data from the United Kingdom, sourced from https://coronavirus.data.gov.uk/", "Name": "uk_covid", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "type of area covered: overview, nation, region, nhsRegion, utla, ltla", "Name": "areatype", "Type": "string" }, { "Comment": "name of the covered area", "Name": "areaname", "Type": "string" }, { "Comment": "code of the covered area", "Name": "areacode", "Type": "string" }, { "Comment": "reporting date", "Name": "date", "Type": "string" }, { "Comment": "New cases by publish date", "Name": "newcasesbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative cases by publish date", "Name": "cumcasesbypublishdate", "Type": "bigint" }, { "Comment": "New cases by specimen date", "Name": "newcasesbyspecimendate", "Type": "bigint" }, { "Comment": "Cumulative cases by specimen date", "Name": "cumcasesbyspecimendate", "Type": "bigint" }, { "Comment": "Male cases (by age)", "Name": "malecases", "Type": "array>" }, { "Comment": "Female cases (by age)", "Name": "femalecases", "Type": "array>" }, { "Comment": "New pillar one tests by publish date", "Name": "newpillaronetestsbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative pillar one tests by publish date", "Name": "cumpillaronetestsbypublishdate", "Type": "bigint" }, { "Comment": "New pillar two tests by publish date", "Name": "newpillartwotestsbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative pillar two tests by publish date", "Name": "cumpillartwotestsbypublishdate", "Type": "bigint" }, { "Comment": "New pillar three tests by publish date", "Name": "newpillarthreetestsbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative pillar three tests by publish date", "Name": "cumpillarthreetestsbypublishdate", "Type": "bigint" }, { "Comment": "New pillar four tests by publish date", "Name": "newpillarfourtestsbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative pillar four tests by publish date", "Name": "cumpillarfourtestsbypublishdate", "Type": "bigint" }, { "Comment": "New admissions", "Name": "newadmissions", "Type": "bigint" }, { "Comment": "Cumulative number of admissions", "Name": "cumadmissions", "Type": "bigint" }, { "Comment": "Cumulative admissions by age", "Name": "cumadmissionsbyage", "Type": "array>" }, { "Comment": "Cumulative tests by publish date", "Name": "cumtestsbypublishdate", "Type": "bigint" }, { "Comment": "New tests by publish date", "Name": "newtestsbypublishdate", "Type": "bigint" }, { "Comment": "COVID-19 occupied beds with mechanical ventilators", "Name": "covidoccupiedmvbeds", "Type": "bigint" }, { "Comment": "Hospital cases", "Name": "hospitalcases", "Type": "bigint" }, { "Comment": "Planned capacity by publish date", "Name": "plannedcapacitybypublishdate", "Type": "bigint" }, { "Comment": "Deaths within 28 days of positive test", "Name": "newdeaths28daysbypublishdate", "Type": "bigint" }, { "Comment": "", "Name": "cumdeaths28daysbypublishdate", "Type": "bigint" }, { "Comment": "Cumulative deaths within 28 days of positive test", "Name": "cumdeaths28daysbypublishdaterate", "Type": "double" }, { "Comment": "Deaths within 28 days of positive test by death date", "Name": "newdeaths28daysbydeathdate", "Type": "bigint" }, { "Comment": "Cumulative deaths within 28 days of positive test by death date", "Name": "cumdeaths28daysbydeathdate", "Type": "bigint" }, { "Comment": "Rate of cumulative deaths within 28 days of positive test per 100k resident population", "Name": "cumdeaths28daysbydeathdaterate", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/uk_covid/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "areaCode,areaName,areaType,covidOccupiedMVBeds,cumAdmissions,cumAdmissionsByAge,cumCasesByPublishDate,cumCasesBySpecimenDate,cumDeaths28DaysByDeathDate,cumDeaths28DaysByDeathDateRate,cumDeaths28DaysByPublishDate,cumDeaths28DaysByPublishDateRate,cumPillarFourTestsByPublishDate,cumPillarOneTestsByPublishDate,cumPillarThreeTestsByPublishDate,cumPillarTwoTestsByPublishDate,cumTestsByPublishDate,date,femaleCases,hospitalCases,maleCases,newAdmissions,newCasesByPublishDate,newCasesBySpecimenDate,newDeaths28DaysByDeathDate,newDeaths28DaysByPublishDate,newPillarFourTestsByPublishDate,newPillarOneTestsByPublishDate,newPillarThreeTestsByPublishDate,newPillarTwoTestsByPublishDate,newTestsByPublishDate,plannedCapacityByPublishDate" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/uk_covid" } }, "cdcvaccinespfizer": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Vaccine distribution data by state for the Pfizer/BioNTech vaccine, sourced from the CDC.", "Name": "cdc_pfizer_vaccine_distribution", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "US State", "Name": "jurisdiction", "Type": "string" }, { "Comment": "Monday of the week in which allocations are provided", "Name": "week_of_allocations", "Type": "string" }, { "Comment": "allocations for first vaccine dose", "Name": "first_dose_allocations", "Type": "bigint" }, { "Comment": "allocations for second vaccine dose", "Name": "second_dose_allocations", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/cdc-pfizer-vaccine-distribution/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "first_dose_allocations,second_dose_allocations,jurisdiction,week_of_allocations" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/cdc_vaccines_pfizer" } }, "cdcvaccinesmoderna": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Vaccine distribution data by state for the Moderna vaccine, sourced from the CDC.", "Name": "cdc_moderna_vaccine_distribution", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "US State", "Name": "jurisdiction", "Type": "string" }, { "Comment": "Monday of the week in which allocations are provided", "Name": "week_of_allocations", "Type": "string" }, { "Comment": "allocations for first vaccine dose", "Name": "first_dose_allocations", "Type": "bigint" }, { "Comment": "allocations for second vaccine dose", "Name": "second_dose_allocations", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/cdc-moderna-vaccine-distribution/json/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "first_dose_allocations,second_dose_allocations,jurisdiction,week_of_allocations" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/cdc_vaccines_moderna" } }, "owidworldvaccinations": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Country-by-country data on global COVID-19 vaccinations, sourced from Our World in Data.", "Name": "owid_world_vaccinations", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "name of the country (or region within a country)", "Name": "country", "Type": "string" }, { "Comment": "ISO 3166-1 alpha-3 – three-letter country codes", "Name": "iso_code", "Type": "string" }, { "Comment": "date of the observation", "Name": "date", "Type": "string" }, { "Comment": "total number of doses administered. This is counted as a single dose, and may not equal the total number of people vaccinated, depending on the specific dose regime.", "Name": "total_vaccinations", "Type": "bigint" }, { "Comment": "total_vaccinations per 100 people in the total population of the country", "Name": "total_vaccinations_per_hundred", "Type": "double" }, { "Comment": "daily change in the total number of doses administered. It is only calculated for consecutive days. This is a raw measure provided for data checks and transparency.", "Name": "daily_vaccinations_raw", "Type": "bigint" }, { "Comment": "new doses administered per day (7-day smoothed). For countries that don't report data on a daily basis, we assume that doses changed equally on a daily basis over any periods in which no data was reported.", "Name": "daily_vaccinations", "Type": "bigint" }, { "Comment": "daily_vaccinations per 1,000,000 people in the total population of the country", "Name": "daily_vaccinations_per_million", "Type": "double" }, { "Comment": "total number of people who received at least one vaccine dose. ", "Name": "people_vaccinated", "Type": "bigint" }, { "Comment": "people_vaccinated per 100 people in the total population of the country", "Name": "people_vaccinated_per_hundred", "Type": "double" }, { "Comment": "total number of people who received all doses prescribed by the vaccination protocol", "Name": "people_fully_vaccinated", "Type": "bigint" }, { "Comment": "people_fully_vaccinated per 100 people in the total population of the country", "Name": "people_fully_vaccinated_per_hundred", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/owid_vaccinations/json/world_vaccinations/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "country,daily_vaccinations,daily_vaccinations_per_million,date,iso_code,people_vaccinated,people_vaccinated_per_hundred,total_vaccinations,total_vaccinations_per_hundred" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/owid_world_vaccinations" } }, "owidusstatevaccinations": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "US data on COVID-19 vaccinations, sourced from Our World in Data.", "Name": "owid_us_state_vaccinations", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "name of the state or federal entity", "Name": "location", "Type": "string" }, { "Comment": "date of the observation", "Name": "date", "Type": "string" }, { "Comment": "total number of doses administered. This is counted as a single dose, and may not equal the total number of people vaccinated, depending on the specific dose regime (e.g. people receive multiple doses).", "Name": "total_vaccinations", "Type": "bigint" }, { "Comment": "total_vaccinations per 100 people in the total population of the state", "Name": "total_vaccinations_per_hundred", "Type": "double" }, { "Comment": "daily change in the total number of doses administered. It is only calculated for consecutive days. This is a raw measure provided for data checks and transparency.", "Name": "daily_vaccinations_raw", "Type": "bigint" }, { "Comment": "new doses administered per day (7-day smoothed). For countries that don't report data on a daily basis, we assume that doses changed equally on a daily basis over any periods in which no data was reported.", "Name": "daily_vaccinations", "Type": "bigint" }, { "Comment": "daily_vaccinations per 1,000,000 people in the total population of the state", "Name": "daily_vaccinations_per_million", "Type": "double" }, { "Comment": "total number of people who received at least one vaccine dose", "Name": "people_vaccinated", "Type": "bigint" }, { "Comment": "people_vaccinated per 100 people in the total population of the state", "Name": "people_vaccinated_per_hundred", "Type": "double" }, { "Comment": "total number of people who received all doses prescribed by the vaccination protocol", "Name": "people_fully_vaccinated", "Type": "bigint" }, { "Comment": "people_fully_vaccinated per 100 people in the total population of the state", "Name": "people_fully_vaccinated_per_hundred", "Type": "double" }, { "Comment": "cumulative counts of COVID-19 vaccine doses recorded as shipped in CDC's Vaccine Tracking System", "Name": "total_distributed", "Type": "bigint" }, { "Comment": "cumulative counts of COVID-19 vaccine doses recorded as shipped in CDC's Vaccine Tracking System per 100 people in the total population of the state", "Name": "total_distributed_per_hundred", "Type": "double" }, { "Comment": "share of vaccination doses administered among those recorded as shipped in CDC's Vaccine Tracking System", "Name": "share_doses_used", "Type": "double" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/owid_vaccinations/json/us_state_vaccinations/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "daily_vaccinations,daily_vaccinations_per_million,daily_vaccinations_raw,date,location,people_fully_vaccinated,people_fully_vaccinated_per_hundred,people_vaccinated,people_vaccinated_per_hundred,share_doses_used,total_distributed,total_distributed_per_hundred,total_vaccinations,total_vaccinations_per_hundred" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/owid_us_state_vaccinations" } }, "owidworldvaccinationsbymanufacturer": { "Type": "AWS::Glue::Table", "Properties": { "CatalogId": { "Ref": "AWS::AccountId" }, "DatabaseName": { "Ref": "covid19AE266CCE" }, "TableInput": { "Description": "Country-by-country data on global COVID-19 vaccinations broken down by vaccine manufacturer, sourced from Our World in Data.", "Name": "owid_world_vaccinations_by_manufacturer", "Parameters": { "has_encrypted_data": false, "classification": "json", "typeOfData": "file" }, "StorageDescriptor": { "Columns": [ { "Comment": "name of the country (or region within a country)", "Name": "location", "Type": "string" }, { "Comment": "date of the observation", "Name": "date", "Type": "string" }, { "Comment": "vaccine manufacturer", "Name": "vaccine", "Type": "string" }, { "Comment": "total number of doses adminisetered", "Name": "total_vaccinations", "Type": "bigint" } ], "Compressed": false, "InputFormat": "org.apache.hadoop.mapred.TextInputFormat", "Location": "s3://covid19-lake/owid_vaccinations/json/world_vaccinations_by_manufacturer/", "OutputFormat": "org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat", "SerdeInfo": { "Parameters": { "paths": "date,location,total_vaccinations,vaccine" }, "SerializationLibrary": "org.openx.data.jsonserde.JsonSerDe" }, "StoredAsSubDirectories": false }, "TableType": "EXTERNAL_TABLE" } }, "Metadata": { "aws:cdk:path": "CovidLakeStack/owid_world_vaccinations_by_manufacturer" } }, "CDKMetadata": { "Type": "AWS::CDK::Metadata", "Properties": { "Modules": "aws-cdk=1.63.0,@aws-cdk/aws-events=1.63.0,@aws-cdk/aws-glue=1.63.0,@aws-cdk/aws-iam=1.63.0,@aws-cdk/aws-kms=1.63.0,@aws-cdk/aws-s3=1.63.0,@aws-cdk/cloud-assembly-schema=1.63.0,@aws-cdk/core=1.63.0,@aws-cdk/cx-api=1.63.0,@aws-cdk/region-info=1.63.0,jsii-runtime=node.js/v12.14.0" }, "Condition": "CDKMetadataAvailable" } }, "Conditions": { "CDKMetadataAvailable": { "Fn::Or": [ { "Fn::Or": [ { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-east-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-northeast-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-northeast-2" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-south-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-southeast-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ap-southeast-2" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "ca-central-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "cn-north-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "cn-northwest-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "eu-central-1" ] } ] }, { "Fn::Or": [ { "Fn::Equals": [ { "Ref": "AWS::Region" }, "eu-north-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "eu-west-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "eu-west-2" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "eu-west-3" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "me-south-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "sa-east-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "us-east-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "us-east-2" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "us-west-1" ] }, { "Fn::Equals": [ { "Ref": "AWS::Region" }, "us-west-2" ] } ] } ] } } }