-
Notifications
You must be signed in to change notification settings - Fork 13
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Create NTD external tables for 2022 API data (#3465)
[#3403] Co-authored-by: Erika Pacheco <[email protected]>
- Loading branch information
Showing
34 changed files
with
2,318 additions
and
0 deletions.
There are no files selected for viewing
58 changes: 58 additions & 0 deletions
58
airflow/dags/create_external_tables/ntd_data_products/2022__breakdowns.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,58 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "breakdowns/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "breakdowns/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__breakdowns" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__breakdowns LIMIT 1; | ||
| schema_fields: | ||
| - name: agency | ||
| type: STRING | ||
| - name: agency_voms | ||
| type: NUMERIC | ||
| - name: city | ||
| type: STRING | ||
| - name: major_mechanical_failures | ||
| type: NUMERIC | ||
| - name: mode | ||
| type: STRING | ||
| - name: mode_name | ||
| type: STRING | ||
| - name: mode_voms | ||
| type: NUMERIC | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: organization_type | ||
| type: STRING | ||
| - name: other_mechanical_failures | ||
| type: NUMERIC | ||
| - name: primary_uza_population | ||
| type: NUMERIC | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: reporter_type | ||
| type: STRING | ||
| - name: state | ||
| type: STRING | ||
| - name: total_mechanical_failures | ||
| type: NUMERIC | ||
| - name: train_miles | ||
| type: NUMERIC | ||
| - name: train_revenue_miles | ||
| type: NUMERIC | ||
| - name: type_of_service | ||
| type: STRING | ||
| - name: uace_code | ||
| type: STRING | ||
| - name: uza_name | ||
| type: STRING | ||
| - name: vehicle_passenger_car_miles | ||
| type: NUMERIC | ||
| - name: vehicle_passenger_car_revenue | ||
| type: NUMERIC |
62 changes: 62 additions & 0 deletions
62
airflow/dags/create_external_tables/ntd_data_products/2022__breakdowns_by_agency.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,62 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "breakdowns_by_agency/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "breakdowns_by_agency/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__breakdowns_by_agency" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__breakdowns_by_agency LIMIT 1; | ||
| schema_fields: | ||
| - name: count_major_mechanical_failures_questionable | ||
| type: NUMERIC | ||
| - name: count_other_mechanical_failures_questionable | ||
| type: NUMERIC | ||
| - name: count_total_mechanical_failures_questionable | ||
| type: NUMERIC | ||
| - name: count_train_miles_questionable | ||
| type: NUMERIC | ||
| - name: count_train_revenue_miles_questionable | ||
| type: NUMERIC | ||
| - name: count_vehicle_passenger_car_miles_questionable | ||
| type: NUMERIC | ||
| - name: max_agency | ||
| type: STRING | ||
| - name: max_agency_voms | ||
| type: NUMERIC | ||
| - name: max_city | ||
| type: STRING | ||
| - name: max_organization_type | ||
| type: STRING | ||
| - name: max_primary_uza_population | ||
| type: NUMERIC | ||
| - name: max_reporter_type | ||
| type: STRING | ||
| - name: max_state | ||
| type: STRING | ||
| - name: max_uace_code | ||
| type: STRING | ||
| - name: max_uza_name | ||
| type: STRING | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: sum_major_mechanical_failures | ||
| type: NUMERIC | ||
| - name: sum_other_mechanical_failures | ||
| type: NUMERIC | ||
| - name: sum_total_mechanical_failures | ||
| type: NUMERIC | ||
| - name: sum_train_miles | ||
| type: NUMERIC | ||
| - name: sum_train_revenue_miles | ||
| type: NUMERIC | ||
| - name: sum_vehicle_passenger_car_miles | ||
| type: NUMERIC | ||
| - name: sum_vehicle_passenger_car_revenue | ||
| type: NUMERIC |
62 changes: 62 additions & 0 deletions
62
...w/dags/create_external_tables/ntd_data_products/2022__capital_expenses_by_capital_use.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,62 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "capital_expenses_by_capital_use/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "capital_expenses_by_capital_use/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__capital_expenses_by_capital_use" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__capital_expenses_by_capital_use LIMIT 1; | ||
| schema_fields: | ||
| - name: administrative_buildings | ||
| type: STRING | ||
| - name: agency | ||
| type: STRING | ||
| - name: agency_voms | ||
| type: NUMERIC | ||
| - name: city | ||
| type: STRING | ||
| - name: communication_information | ||
| type: STRING | ||
| - name: fare_collection_equipment | ||
| type: STRING | ||
| - name: form_type | ||
| type: STRING | ||
| - name: guideway | ||
| type: STRING | ||
| - name: maintenance_buildings | ||
| type: STRING | ||
| - name: mode_name | ||
| type: STRING | ||
| - name: mode_voms | ||
| type: NUMERIC | ||
| - name: modecd | ||
| type: STRING | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: organization_type | ||
| type: STRING | ||
| - name: other | ||
| type: STRING | ||
| - name: other_vehicles | ||
| type: STRING | ||
| - name: passenger_vehicles | ||
| type: STRING | ||
| - name: reduced_reporter | ||
| type: STRING | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: reporter_type | ||
| type: STRING | ||
| - name: state | ||
| type: STRING | ||
| - name: stations | ||
| type: STRING | ||
| - name: total | ||
| type: NUMERIC | ||
| - name: typeofservicecd | ||
| type: STRING |
76 changes: 76 additions & 0 deletions
76
airflow/dags/create_external_tables/ntd_data_products/2022__capital_expenses_by_mode.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,76 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "capital_expenses_by_mode/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "capital_expenses_by_mode/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__capital_expenses_by_mode" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__capital_expenses_by_mode LIMIT 1; | ||
| schema_fields: | ||
| - name: count_administrative_buildings_q | ||
| type: NUMERIC | ||
| - name: count_communication_information_q | ||
| type: NUMERIC | ||
| - name: count_fare_collection_equipment_q | ||
| type: NUMERIC | ||
| - name: count_maintenance_buildings_q | ||
| type: NUMERIC | ||
| - name: count_other_q | ||
| type: NUMERIC | ||
| - name: count_other_vehicles_q | ||
| type: NUMERIC | ||
| - name: count_passenger_vehicles_q | ||
| type: NUMERIC | ||
| - name: count_reduced_reporter_q | ||
| type: NUMERIC | ||
| - name: count_stations_q | ||
| type: NUMERIC | ||
| - name: max_agency | ||
| type: STRING | ||
| - name: max_agency_voms | ||
| type: STRING | ||
| - name: max_city | ||
| type: STRING | ||
| - name: max_mode_name | ||
| type: STRING | ||
| - name: max_organization_type | ||
| type: STRING | ||
| - name: max_reporter_type | ||
| type: STRING | ||
| - name: max_state | ||
| type: STRING | ||
| - name: modecd | ||
| type: STRING | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: sum_administrative_buildings | ||
| type: NUMERIC | ||
| - name: sum_communication_information | ||
| type: NUMERIC | ||
| - name: sum_fare_collection_equipment | ||
| type: NUMERIC | ||
| - name: sum_guideway | ||
| type: NUMERIC | ||
| - name: sum_maintenance_buildings | ||
| type: NUMERIC | ||
| - name: sum_other | ||
| type: NUMERIC | ||
| - name: sum_other_vehicles | ||
| type: NUMERIC | ||
| - name: sum_passenger_vehicles | ||
| type: NUMERIC | ||
| - name: sum_reduced_reporter | ||
| type: NUMERIC | ||
| - name: sum_stations | ||
| type: NUMERIC | ||
| - name: sum_total | ||
| type: NUMERIC | ||
| - name: typeofservicecd | ||
| type: STRING |
60 changes: 60 additions & 0 deletions
60
.../create_external_tables/ntd_data_products/2022__capital_expenses_for_existing_service.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,60 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "capital_expenses_for_existing_service/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "capital_expenses_for_existing_service/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__capital_expenses_for_existing_service" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__capital_expenses_for_existing_service LIMIT 1; | ||
| schema_fields: | ||
| - name: form_type | ||
| type: STRING | ||
| - name: max_agency | ||
| type: STRING | ||
| - name: max_agency_voms | ||
| type: STRING | ||
| - name: max_city | ||
| type: STRING | ||
| - name: max_organization_type | ||
| type: STRING | ||
| - name: max_primary_uza_population | ||
| type: STRING | ||
| - name: max_reporter_type | ||
| type: STRING | ||
| - name: max_state | ||
| type: STRING | ||
| - name: max_uace_code | ||
| type: STRING | ||
| - name: max_uza_name | ||
| type: STRING | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: sum_administrative_buildings | ||
| type: NUMERIC | ||
| - name: sum_communication_information | ||
| type: NUMERIC | ||
| - name: sum_fare_collection_equipment | ||
| type: NUMERIC | ||
| - name: sum_guideway | ||
| type: NUMERIC | ||
| - name: sum_maintenance_buildings | ||
| type: NUMERIC | ||
| - name: sum_other | ||
| type: NUMERIC | ||
| - name: sum_other_vehicles | ||
| type: NUMERIC | ||
| - name: sum_passenger_vehicles | ||
| type: NUMERIC | ||
| - name: sum_reduced_reporter | ||
| type: NUMERIC | ||
| - name: sum_stations | ||
| type: NUMERIC | ||
| - name: sum_total | ||
| type: NUMERIC |
60 changes: 60 additions & 0 deletions
60
...ate_external_tables/ntd_data_products/2022__capital_expenses_for_expansion_of_service.yml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,60 @@ | ||
| operator: operators.ExternalTable | ||
| bucket: gs://calitp-ntd-api-products | ||
| source_objects: | ||
| - "capital_expenses_for_expansion_of_service/2022/*.jsonl.gz" | ||
| source_format: NEWLINE_DELIMITED_JSON | ||
| use_bq_client: true | ||
| hive_options: | ||
| mode: CUSTOM | ||
| require_partition_filter: false | ||
| source_uri_prefix: "capital_expenses_for_expansion_of_service/2022/{dt:DATE}/{execution_ts:TIMESTAMP}" | ||
| destination_project_dataset_table: "external_ntd__annual_reporting.2022__capital_expenses_for_expansion_of_service" | ||
| prefix_bucket: false | ||
| post_hook: SELECT * FROM `{{ get_project_id() }}`.external_ntd__annual_reporting.2022__capital_expenses_for_expansion_of_service LIMIT 1; | ||
| schema_fields: | ||
| - name: form_type | ||
| type: STRING | ||
| - name: max_agency | ||
| type: STRING | ||
| - name: max_agency_voms | ||
| type: STRING | ||
| - name: max_city | ||
| type: STRING | ||
| - name: max_organization_type | ||
| type: STRING | ||
| - name: max_primary_uza_population | ||
| type: NUMERIC | ||
| - name: max_reporter_type | ||
| type: STRING | ||
| - name: max_state | ||
| type: STRING | ||
| - name: max_uace_code | ||
| type: STRING | ||
| - name: max_uza_name | ||
| type: STRING | ||
| - name: ntd_id | ||
| type: NUMERIC | ||
| - name: report_year | ||
| type: NUMERIC | ||
| - name: sum_administrative_buildings | ||
| type: NUMERIC | ||
| - name: sum_communication_information | ||
| type: NUMERIC | ||
| - name: sum_fare_collection_equipment | ||
| type: NUMERIC | ||
| - name: sum_guideway | ||
| type: NUMERIC | ||
| - name: sum_maintenance_buildings | ||
| type: NUMERIC | ||
| - name: sum_other | ||
| type: NUMERIC | ||
| - name: sum_other_vehicles | ||
| type: NUMERIC | ||
| - name: sum_passenger_vehicles | ||
| type: NUMERIC | ||
| - name: sum_reduced_reporter | ||
| type: NUMERIC | ||
| - name: sum_stations | ||
| type: NUMERIC | ||
| - name: sum_total | ||
| type: NUMERIC |
Oops, something went wrong.