From 3da2004ce7839d96d1279c1d817f0d41e070280a Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Thu, 9 Sep 2021 08:05:43 +0000 Subject: [PATCH 1/8] feat: Onboard Chicago Crime dataset --- .../_images/run_csv_transform_kub/Dockerfile | 38 +++ .../run_csv_transform_kub/csv_transform.py | 222 ++++++++++++++++++ .../run_csv_transform_kub/requirements.txt | 3 + .../_terraform/chicago_crime_dataset.tf | 26 ++ .../_terraform/crime_pipeline.tf | 39 +++ datasets/chicago_crime/_terraform/provider.tf | 28 +++ .../chicago_crime/_terraform/terraform.tfvars | 23 ++ .../chicago_crime/_terraform/variables.tf | 23 ++ datasets/chicago_crime/crime/crime_dag.py | 88 +++++++ datasets/chicago_crime/crime/pipeline.yaml | 169 +++++++++++++ datasets/chicago_crime/dataset.yaml | 67 ++++++ 11 files changed, 726 insertions(+) create mode 100644 datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile create mode 100644 datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py create mode 100644 datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt create mode 100644 datasets/chicago_crime/_terraform/chicago_crime_dataset.tf create mode 100644 datasets/chicago_crime/_terraform/crime_pipeline.tf create mode 100644 datasets/chicago_crime/_terraform/provider.tf create mode 100644 datasets/chicago_crime/_terraform/terraform.tfvars create mode 100644 datasets/chicago_crime/_terraform/variables.tf create mode 100644 datasets/chicago_crime/crime/crime_dag.py create mode 100644 datasets/chicago_crime/crime/pipeline.yaml create mode 100644 datasets/chicago_crime/dataset.yaml diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile b/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile new file mode 100644 index 000000000..85af90570 --- /dev/null +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile @@ -0,0 +1,38 @@ +# Copyright 2021 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +# The base image for this build +# FROM gcr.io/google.com/cloudsdktool/cloud-sdk:slim +FROM python:3.8 + +# Allow statements and log messages to appear in Cloud logs +ENV PYTHONUNBUFFERED True + +# Copy the requirements file into the image +COPY requirements.txt ./ + +# Install the packages specified in the requirements file +RUN python3 -m pip install --no-cache-dir -r requirements.txt + +# The WORKDIR instruction sets the working directory for any RUN, CMD, +# ENTRYPOINT, COPY and ADD instructions that follow it in the Dockerfile. +# If the WORKDIR doesn’t exist, it will be created even if it’s not used in +# any subsequent Dockerfile instruction +WORKDIR /custom + +# Copy the specific data processing script/s in the image under /custom/* +COPY ./csv_transform.py . + +# Command to run the data processing script when the container is run +CMD ["python3", "csv_transform.py"] diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py new file mode 100644 index 000000000..3d769fc3c --- /dev/null +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py @@ -0,0 +1,222 @@ +# Copyright 2021 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + + +import datetime +import logging +import math +import os +import pathlib +import typing + +import pandas as pd +import requests +from google.cloud import storage + + +def main( + source_url: str, + source_file: pathlib.Path, + target_file: pathlib.Path, + target_gcs_bucket: str, + target_gcs_path: str, +) -> None: + + logging.info( + "chicago crime process started at " + + str(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")) + ) + + logging.info("creating 'files' folder") + pathlib.Path("./files").mkdir(parents=True, exist_ok=True) + + logging.info(f"Downloading file {source_url}") + download_file(source_url, source_file) + + logging.info(f"Opening file {source_file}") + df = pd.read_csv(source_file) + + logging.info(f"Transforming {source_file} ...") + + logging.info(f"Transform: Rename columns {source_file} ...") + rename_headers(df) + + logging.info("Transform: Converting date format.. ") + convert_values(df) + + logging.info("Transform: Removing null values.. ") + filter_null_rows(df) + + logging.info("Transform: Reordering headers..") + df = df[ + [ + "unique_key", + "case_number", + "date", + "block", + "iucr", + "primary_type", + "description", + "location_description", + "arrest", + "domestic", + "beat", + "district", + "ward", + "community_area", + "fbi_code", + "x_coordinate", + "y_coordinate", + "year", + "updated_on", + "latitude", + "longitude", + "location", + ] + ] + + logging.info("Transform: converting to integers..") + + df["unique_key"] = df["unique_key"].apply(convert_to_integer_string) + df["beat"] = df["beat"].apply(convert_to_integer_string) + df["district"] = df["district"].apply(convert_to_integer_string) + df["ward"] = df["ward"].apply(convert_to_integer_string) + df["community_area"] = df["community_area"].apply(convert_to_integer_string) + df["year"] = df["year"].apply(convert_to_integer_string) + + logging.info("Transform: converting to float..") + + df["x_coordinate"] = df["x_coordinate"].apply(resolve_nan) + df["y_coordinate"] = df["y_coordinate"].apply(resolve_nan) + df["latitude"] = df["latitude"].apply(resolve_nan) + df["longitude"] = df["longitude"].apply(resolve_nan) + + logging.info(f"Saving to output file.. {target_file}") + try: + save_to_new_file(df, file_path=str(target_file)) + except Exception as e: + logging.error(f"Error saving output file: {e}.") + + logging.info( + f"Uploading output file to.. gs://{target_gcs_bucket}/{target_gcs_path}" + ) + upload_file_to_gcs(target_file, target_gcs_bucket, target_gcs_path) + + logging.info( + "Chicago crime process completed at " + + str(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")) + ) + + +def resolve_nan(input: typing.Union[str, float]) -> str: + str_val = "" + if not input or (math.isnan(input)): + str_val = "" + else: + str_val = str(input) + return str_val.replace("None", "") + + +def convert_to_integer_string(input: typing.Union[str, float]) -> str: + str_val = "" + if not input or (math.isnan(input)): + str_val = "" + else: + str_val = str(int(round(input, 0))) + return str_val + + +def rename_headers(df: pd.DataFrame) -> None: + header_names = { + "ID": "unique_key", + "Case Number": "case_number", + "Date": "date", + "Block": "block", + "IUCR": "iucr", + "Primary Type": "primary_type", + "Description": "description", + "Location Description": "location_description", + "Arrest": "arrest", + "Domestic": "domestic", + "Beat": "beat", + "District": "district", + "Ward": "ward", + "Community Area": "community_area", + "FBI Code": "fbi_code", + "X Coordinate": "x_coordinate", + "Y Coordinate": "y_coordinate", + "Year": "year", + "Updated On": "updated_on", + "Latitude": "latitude", + "Longitude": "longitude", + "Location": "location", + } + + df.rename(columns=header_names, inplace=True) + + +def convert_dt_format(dt_str: str) -> str: + # Old format: MM/dd/yyyy hh:mm:ss aa + # New format: yyyy-MM-dd HH:mm:ss + if dt_str is None or len(dt_str) == 0: + return dt_str + else: + return datetime.datetime.strptime(dt_str, "%m/%d/%Y %H:%M:%S %p").strftime( + "%Y-%m-%d %H:%M:%S" + ) + + +def convert_values(df: pd.DataFrame) -> None: + dt_cols = ["date", "updated_on"] + + for dt_col in dt_cols: + df[dt_col] = df[dt_col].apply(convert_dt_format) + + +def filter_null_rows(df: pd.DataFrame) -> None: + df = df[df.unique_key != ""] + + +def save_to_new_file(df: pd.DataFrame, file_path: pathlib.Path) -> None: + df.to_csv(file_path, index=False) + + +def download_file(source_url: str, source_file: pathlib.Path) -> None: + logging.info(f"Downloading {source_url} into {source_file}") + r = requests.get(source_url, stream=True) + if r.status_code == 200: + with open(source_file, "wb") as f: + for chunk in r: + f.write(chunk) + else: + logging.error(f"Couldn't download {source_url}: {r.text}") + + +def upload_file_to_gcs(file_path: pathlib.Path, gcs_bucket: str, gcs_path: str) -> None: + storage_client = storage.Client() + bucket = storage_client.bucket(gcs_bucket) + blob = bucket.blob(gcs_path) + blob.upload_from_filename(file_path) + + +if __name__ == "__main__": + logging.getLogger().setLevel(logging.INFO) + + main( + source_url=os.environ["SOURCE_URL"], + source_file=pathlib.Path(os.environ["SOURCE_FILE"]).expanduser(), + target_file=pathlib.Path(os.environ["TARGET_FILE"]).expanduser(), + target_gcs_bucket=os.environ["TARGET_GCS_BUCKET"], + target_gcs_path=os.environ["TARGET_GCS_PATH"], + ) diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt b/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt new file mode 100644 index 000000000..64755a48d --- /dev/null +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt @@ -0,0 +1,3 @@ +requests +vaex +google-cloud-storage diff --git a/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf b/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf new file mode 100644 index 000000000..db644febe --- /dev/null +++ b/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf @@ -0,0 +1,26 @@ +/** + * Copyright 2021 Google LLC + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + + +resource "google_bigquery_dataset" "chicago_crime" { + dataset_id = "chicago_crime" + project = var.project_id + description = "chicago crime" +} + +output "bigquery_dataset-chicago_crime-dataset_id" { + value = google_bigquery_dataset.chicago_crime.dataset_id +} diff --git a/datasets/chicago_crime/_terraform/crime_pipeline.tf b/datasets/chicago_crime/_terraform/crime_pipeline.tf new file mode 100644 index 000000000..ab7f09462 --- /dev/null +++ b/datasets/chicago_crime/_terraform/crime_pipeline.tf @@ -0,0 +1,39 @@ +/** + * Copyright 2021 Google LLC + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + + +resource "google_bigquery_table" "crime" { + project = var.project_id + dataset_id = "chicago_crime" + table_id = "crime" + + description = "chicago crime dataset" + + + + + depends_on = [ + google_bigquery_dataset.chicago_crime + ] +} + +output "bigquery_table-crime-table_id" { + value = google_bigquery_table.crime.table_id +} + +output "bigquery_table-crime-id" { + value = google_bigquery_table.crime.id +} diff --git a/datasets/chicago_crime/_terraform/provider.tf b/datasets/chicago_crime/_terraform/provider.tf new file mode 100644 index 000000000..23ab87dcd --- /dev/null +++ b/datasets/chicago_crime/_terraform/provider.tf @@ -0,0 +1,28 @@ +/** + * Copyright 2021 Google LLC + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + + +provider "google" { + project = var.project_id + impersonate_service_account = var.impersonating_acct + region = var.region +} + +data "google_client_openid_userinfo" "me" {} + +output "impersonating-account" { + value = data.google_client_openid_userinfo.me.email +} diff --git a/datasets/chicago_crime/_terraform/terraform.tfvars b/datasets/chicago_crime/_terraform/terraform.tfvars new file mode 100644 index 000000000..1ce7806c8 --- /dev/null +++ b/datasets/chicago_crime/_terraform/terraform.tfvars @@ -0,0 +1,23 @@ +/** + * Copyright 2021 Google LLC + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + + +project_id = "bigquery-public-data-dev" +bucket_name_prefix = "public-datasets-dev" +impersonating_acct = "datasets-pipelines-dev-tf@bigquery-public-data-dev.iam.gserviceaccount.com" +region = "us-central1" +env = "dev" + diff --git a/datasets/chicago_crime/_terraform/variables.tf b/datasets/chicago_crime/_terraform/variables.tf new file mode 100644 index 000000000..c3ec7c506 --- /dev/null +++ b/datasets/chicago_crime/_terraform/variables.tf @@ -0,0 +1,23 @@ +/** + * Copyright 2021 Google LLC + * + * Licensed under the Apache License, Version 2.0 (the "License"); + * you may not use this file except in compliance with the License. + * You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + + +variable "project_id" {} +variable "bucket_name_prefix" {} +variable "impersonating_acct" {} +variable "region" {} +variable "env" {} + diff --git a/datasets/chicago_crime/crime/crime_dag.py b/datasets/chicago_crime/crime/crime_dag.py new file mode 100644 index 000000000..b0226dd0b --- /dev/null +++ b/datasets/chicago_crime/crime/crime_dag.py @@ -0,0 +1,88 @@ +# Copyright 2021 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + + +from airflow import DAG +from airflow.contrib.operators import gcs_to_bq, kubernetes_pod_operator + +default_args = { + "owner": "Google", + "depends_on_past": False, + "start_date": "2021-03-01", +} + + +with DAG( + dag_id="chicago_crime.crime", + default_args=default_args, + max_active_runs=1, + schedule_interval="@daily", + catchup=False, + default_view="graph", +) as dag: + + # Run CSV transform within kubernetes pod + chicago_crime_transform_csv = kubernetes_pod_operator.KubernetesPodOperator( + task_id="chicago_crime_transform_csv", + startup_timeout_seconds=600, + name="crime", + namespace="default", + image_pull_policy="Always", + image="{{ var.json.chicago_crime.container_registry.run_csv_transform_kub }}", + env_vars={ + "SOURCE_URL": "https://data.cityofchicago.org/api/views/ijzp-q8t2/rows.csv", + "SOURCE_FILE": "files/data.csv", + "TARGET_FILE": "files/data_output.csv", + "TARGET_GCS_BUCKET": "{{ var.json.shared.composer_bucket }}", + "TARGET_GCS_PATH": "data/chicago_crime/crime/data_output.csv", + }, + resources={"request_memory": "8G", "request_cpu": "2"}, + ) + + # Task to load CSV data to a BigQuery table + load_chicago_crime_to_bq = gcs_to_bq.GoogleCloudStorageToBigQueryOperator( + task_id="load_chicago_crime_to_bq", + bucket="{{ var.json.shared.composer_bucket }}", + source_objects=["data/chicago_crime/crime/data_output.csv"], + source_format="CSV", + destination_project_dataset_table="chicago_crime.crime", + skip_leading_rows=1, + write_disposition="WRITE_TRUNCATE", + schema_fields=[ + {"name": "unique_key", "type": "integer", "mode": "required"}, + {"name": "case_number", "type": "string", "mode": "nullable"}, + {"name": "date", "type": "timestamp", "mode": "nullable"}, + {"name": "block", "type": "string", "mode": "nullable"}, + {"name": "iucr", "type": "string", "mode": "nullable"}, + {"name": "primary_type", "type": "string", "mode": "nullable"}, + {"name": "description", "type": "string", "mode": "nullable"}, + {"name": "location_description", "type": "string", "mode": "nullable"}, + {"name": "arrest", "type": "boolean", "mode": "nullable"}, + {"name": "domestic", "type": "boolean", "mode": "nullable"}, + {"name": "beat", "type": "integer", "mode": "nullable"}, + {"name": "district", "type": "integer", "mode": "nullable"}, + {"name": "ward", "type": "integer", "mode": "nullable"}, + {"name": "community_area", "type": "integer", "mode": "nullable"}, + {"name": "fbi_code", "type": "string", "mode": "nullable"}, + {"name": "x_coordinate", "type": "float", "mode": "nullable"}, + {"name": "y_coordinate", "type": "float", "mode": "nullable"}, + {"name": "year", "type": "integer"}, + {"name": "updated_on", "type": "timestamp", "mode": "nullable"}, + {"name": "latitude", "type": "float", "mode": "nullable"}, + {"name": "longitude", "type": "float", "mode": "nullable"}, + {"name": "location", "type": "string", "mode": "nullable"}, + ], + ) + + chicago_crime_transform_csv >> load_chicago_crime_to_bq diff --git a/datasets/chicago_crime/crime/pipeline.yaml b/datasets/chicago_crime/crime/pipeline.yaml new file mode 100644 index 000000000..96125b443 --- /dev/null +++ b/datasets/chicago_crime/crime/pipeline.yaml @@ -0,0 +1,169 @@ +# Copyright 2021 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +--- +resources: + + - type: bigquery_table + # Required Properties: + table_id: crime + + # Description of the table + description: "Chicago Crime dataset" + +dag: + airflow_version: 1 + initialize: + dag_id: crime + default_args: + owner: "Google" + + # When set to True, keeps a task from getting triggered if the previous schedule for the task hasn’t succeeded + depends_on_past: False + start_date: '2021-03-01' + max_active_runs: 1 + schedule_interval: "@daily" # runs everyday at 7am EST + catchup: False + default_view: graph + + tasks: + - operator: "KubernetesPodOperator" + + # Task description + description: "Run CSV transform within kubernetes pod" + + args: + + task_id: "chicago_crime_transform_csv" + + startup_timeout_seconds: 600 + + # The name of the pod in which the task will run. This will be used (plus a random suffix) to generate a pod id + name: "crime" + + # The namespace to run within Kubernetes. Always set its value to "default" because we follow the guideline that KubernetesPodOperator will only be used for very light workloads, i.e. use the Cloud Composer environment's resources without starving other pipelines. + namespace: "default" + + image_pull_policy: "Always" + + # Docker images will be built and pushed to GCR by default whenever the `scripts/generate_dag.py` is run. To skip building and pushing images, use the optional `--skip-builds` flag. + image: "{{ var.json.chicago_crime.container_registry.run_csv_transform_kub }}" + + # Set the environment variables you need initialized in the container. Use these as input variables for the script your container is expected to perform. + env_vars: + SOURCE_URL: "https://data.cityofchicago.org/api/views/ijzp-q8t2/rows.csv" + SOURCE_FILE: "files/data.csv" + TARGET_FILE: "files/data_output.csv" + TARGET_GCS_BUCKET: "{{ var.json.shared.composer_bucket }}" + TARGET_GCS_PATH: "data/chicago_crime/crime/data_output.csv" + + # Set resource limits for the pod here. For resource units in Kubernetes, see https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#resource-units-in-kubernetes + resources: + request_memory: "8G" + request_cpu: "2" + + - operator: "GoogleCloudStorageToBigQueryOperator" + description: "Task to load CSV data to a BigQuery table" + + args: + task_id: "load_chicago_crime_to_bq" + + # The GCS bucket where the CSV file is located in. + bucket: "{{ var.json.shared.composer_bucket }}" + + # The GCS object path for the CSV file + source_objects: ["data/chicago_crime/crime/data_output.csv"] + source_format: "CSV" + destination_project_dataset_table: "chicago_crime.crime" + + # Use this if your CSV file contains a header row + skip_leading_rows: 1 + + # How to write data to the table: overwrite, append, or write if empty + # See https://cloud.google.com/bigquery/docs/reference/auditlogs/rest/Shared.Types/WriteDisposition + write_disposition: "WRITE_TRUNCATE" + + # The BigQuery table schema based on the CSV file. For more info, see + # https://cloud.google.com/bigquery/docs/schemas. + # Always use snake_case and lowercase for column names, and be explicit, + # i.e. specify modes for all columns. + schema_fields: + - name: "unique_key" + type: "integer" + mode: "required" + - name: "case_number" + type: "string" + mode: "nullable" + - name: "date" + type: "timestamp" + mode: "nullable" + - name: "block" + type: "string" + mode: "nullable" + - name: "iucr" + type: "string" + mode: "nullable" + - name: "primary_type" + type: "string" + mode: "nullable" + - name: "description" + type: "string" + mode: "nullable" + - name: "location_description" + type: "string" + mode: "nullable" + - name: "arrest" + type: "boolean" + mode: "nullable" + - name: "domestic" + type: "boolean" + mode: "nullable" + - name: "beat" + type: "integer" + mode: "nullable" + - name: "district" + type: "integer" + mode: "nullable" + - name: "ward" + type: "integer" + mode: "nullable" + - name: "community_area" + type: "integer" + mode: "nullable" + - name: "fbi_code" + type: "string" + mode: "nullable" + - name: "x_coordinate" + type: "float" + mode: "nullable" + - name: "y_coordinate" + type: "float" + mode: "nullable" + - name: "year" + type: "integer" + - name: "updated_on" + type: "timestamp" + mode: "nullable" + - name: "latitude" + type: "float" + mode: "nullable" + - name: "longitude" + type: "float" + mode: "nullable" + - name: "location" + type: "string" + mode: "nullable" + + graph_paths: + - "chicago_crime_transform_csv >> load_chicago_crime_to_bq" diff --git a/datasets/chicago_crime/dataset.yaml b/datasets/chicago_crime/dataset.yaml new file mode 100644 index 000000000..fc1110d7f --- /dev/null +++ b/datasets/chicago_crime/dataset.yaml @@ -0,0 +1,67 @@ +# Copyright 2021 Google LLC +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +dataset: + # The `dataset` block includes properties for your dataset that will be shown + # to users of your data on the Google Cloud website. + + # Must be exactly the same name as the folder name your dataset.yaml is in. + name: chicago_crime + + # A friendly, human-readable name of the dataset + friendly_name: chicago_crime + + # A short, descriptive summary of the dataset. + description: |- + This dataset reflects reported incidents of crime (with the exception of murders where data exists for each victim) that occurred in the City of Chicago from 2001 to present, minus the most recent seven days. Data is extracted from the Chicago Police Department's CLEAR (Citizen Law Enforcement Analysis and Reporting) system. In order to protect the privacy of crime victims, addresses are shown at the block level only and specific locations are not identified. This data includes unverified reports supplied to the Police Department. The preliminary crime classifications may be changed at a later date based upon additional investigation and there is always the possibility of mechanical or human error. Therefore, the Chicago Police Department does not guarantee (either expressed or implied) the accuracy, completeness, timeliness, or correct sequencing of the information and the information should not be used for comparison purposes over time. + + Dataset Source: City of Chicago + + Category: Chicago, Public Safety + + Use: This dataset is publicly available for anyone to use under the following terms provided by the Dataset Source —https://data.cityofchicago.org — and is provided "AS IS" without any warranty, express or implied, from Google. Google disclaims all liability for any damages, direct or indirect, resulting from the use of the dataset. + + Update Frequency: Daily + + # A list of sources the dataset is derived from, using the YAML list syntax. + dataset_sources: ~ + + # A list of terms and conditions that users of the dataset should agree on, + # using the YAML list syntax. + terms_of_use: ~ + + +resources: + # A list of Google Cloud resources needed by your dataset. In principle, all + # pipelines under a dataset should be able to share these resources. + # + # The currently supported resources are shown below. Use only the resources + # you need, and delete the rest as needed by your pipeline. + # + # We will keep adding to the list below to support more Google Cloud resources + # over time. If a resource you need isn't supported, please file an issue on + # the repository. + + - type: bigquery_dataset + # Google BigQuery dataset to namespace all tables managed by this folder + # + # Required Properties: + # dataset_id + # + # Optional Properties: + # friendly_name (A user-friendly name of the dataset) + # description (A user-friendly description of the dataset) + # location (The geographic location where the dataset should reside) + dataset_id: chicago_crime + description: chicago crime From 2b649723f1843cdff077e3196df9e3074509dd3c Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Mon, 13 Sep 2021 07:11:59 +0000 Subject: [PATCH 2/8] feat: Onboard Chicago Crime dataset --- datasets/chicago_crime/crime/crime_dag.py | 17 +++++++++++++++++ datasets/chicago_crime/crime/pipeline.yaml | 11 +++++++++++ 2 files changed, 28 insertions(+) diff --git a/datasets/chicago_crime/crime/crime_dag.py b/datasets/chicago_crime/crime/crime_dag.py index b0226dd0b..8de82c093 100644 --- a/datasets/chicago_crime/crime/crime_dag.py +++ b/datasets/chicago_crime/crime/crime_dag.py @@ -38,6 +38,23 @@ startup_timeout_seconds=600, name="crime", namespace="default", + affinity={ + "nodeAffinity": { + "requiredDuringSchedulingIgnoredDuringExecution": { + "nodeSelectorTerms": [ + { + "matchExpressions": [ + { + "key": "cloud.google.com/gke-nodepool", + "operator": "In", + "values": ["pool-e2-standard-4"], + } + ] + } + ] + } + } + }, image_pull_policy="Always", image="{{ var.json.chicago_crime.container_registry.run_csv_transform_kub }}", env_vars={ diff --git a/datasets/chicago_crime/crime/pipeline.yaml b/datasets/chicago_crime/crime/pipeline.yaml index 96125b443..e8c3c7115 100644 --- a/datasets/chicago_crime/crime/pipeline.yaml +++ b/datasets/chicago_crime/crime/pipeline.yaml @@ -55,6 +55,17 @@ dag: # The namespace to run within Kubernetes. Always set its value to "default" because we follow the guideline that KubernetesPodOperator will only be used for very light workloads, i.e. use the Cloud Composer environment's resources without starving other pipelines. namespace: "default" + affinity: + nodeAffinity: + requiredDuringSchedulingIgnoredDuringExecution: + nodeSelectorTerms: + - matchExpressions: + - key: cloud.google.com/gke-nodepool + operator: In + values: + - "pool-e2-standard-4" + + image_pull_policy: "Always" # Docker images will be built and pushed to GCR by default whenever the `scripts/generate_dag.py` is run. To skip building and pushing images, use the optional `--skip-builds` flag. From 05a43aa64706420432c45fed432a41911ba46ee6 Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Fri, 17 Sep 2021 07:00:05 +0000 Subject: [PATCH 3/8] fix: working on review comments --- .../run_csv_transform_kub/csv_transform.py | 38 ++++++++++++++----- .../run_csv_transform_kub/requirements.txt | 2 +- datasets/chicago_crime/crime/pipeline.yaml | 28 +++++--------- 3 files changed, 38 insertions(+), 30 deletions(-) diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py index 3d769fc3c..ff1463a1c 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py @@ -88,19 +88,23 @@ def main( logging.info("Transform: converting to integers..") - df["unique_key"] = df["unique_key"].apply(convert_to_integer_string) - df["beat"] = df["beat"].apply(convert_to_integer_string) - df["district"] = df["district"].apply(convert_to_integer_string) - df["ward"] = df["ward"].apply(convert_to_integer_string) - df["community_area"] = df["community_area"].apply(convert_to_integer_string) - df["year"] = df["year"].apply(convert_to_integer_string) + # df["unique_key"] = df["unique_key"].apply(convert_to_integer_string) + # df["beat"] = df["beat"].apply(convert_to_integer_string) + # df["district"] = df["district"].apply(convert_to_integer_string) + # df["ward"] = df["ward"].apply(convert_to_integer_string) + # df["community_area"] = df["community_area"].apply(convert_to_integer_string) + # df["year"] = df["year"].apply(convert_to_integer_string) + + convert_values_to_integer_string(df) logging.info("Transform: converting to float..") - df["x_coordinate"] = df["x_coordinate"].apply(resolve_nan) - df["y_coordinate"] = df["y_coordinate"].apply(resolve_nan) - df["latitude"] = df["latitude"].apply(resolve_nan) - df["longitude"] = df["longitude"].apply(resolve_nan) + # df["x_coordinate"] = df["x_coordinate"].apply(resolve_nan) + # df["y_coordinate"] = df["y_coordinate"].apply(resolve_nan) + # df["latitude"] = df["latitude"].apply(resolve_nan) + # df["longitude"] = df["longitude"].apply(resolve_nan) + + removing_nan_values(df) logging.info(f"Saving to output file.. {target_file}") try: @@ -128,6 +132,13 @@ def resolve_nan(input: typing.Union[str, float]) -> str: return str_val.replace("None", "") +def removing_nan_values(df: pd.DataFrame) -> None: + cols = ["x_coordinate", "y_coordinate", "latitude", "longitude"] + + for cols in cols: + df[cols] = df[cols].apply(convert_dt_format) + + def convert_to_integer_string(input: typing.Union[str, float]) -> str: str_val = "" if not input or (math.isnan(input)): @@ -137,6 +148,13 @@ def convert_to_integer_string(input: typing.Union[str, float]) -> str: return str_val +def convert_values_to_integer_string(df: pd.DataFrame) -> None: + cols = ["unique_key", "beat", "district", "ward", "community_area", "year"] + + for cols in cols: + df[cols] = df[cols].apply(convert_dt_format) + + def rename_headers(df: pd.DataFrame) -> None: header_names = { "ID": "unique_key", diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt b/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt index 64755a48d..f36704793 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/requirements.txt @@ -1,3 +1,3 @@ requests -vaex +pandas google-cloud-storage diff --git a/datasets/chicago_crime/crime/pipeline.yaml b/datasets/chicago_crime/crime/pipeline.yaml index e8c3c7115..86c382014 100644 --- a/datasets/chicago_crime/crime/pipeline.yaml +++ b/datasets/chicago_crime/crime/pipeline.yaml @@ -16,10 +16,9 @@ resources: - type: bigquery_table - # Required Properties: table_id: crime - # Description of the table + description: "Chicago Crime dataset" dag: @@ -29,18 +28,18 @@ dag: default_args: owner: "Google" - # When set to True, keeps a task from getting triggered if the previous schedule for the task hasn’t succeeded + depends_on_past: False start_date: '2021-03-01' max_active_runs: 1 - schedule_interval: "@daily" # runs everyday at 7am EST + schedule_interval: "@daily" catchup: False default_view: graph tasks: - operator: "KubernetesPodOperator" - # Task description + description: "Run CSV transform within kubernetes pod" args: @@ -49,10 +48,9 @@ dag: startup_timeout_seconds: 600 - # The name of the pod in which the task will run. This will be used (plus a random suffix) to generate a pod id + name: "crime" - # The namespace to run within Kubernetes. Always set its value to "default" because we follow the guideline that KubernetesPodOperator will only be used for very light workloads, i.e. use the Cloud Composer environment's resources without starving other pipelines. namespace: "default" affinity: @@ -68,10 +66,9 @@ dag: image_pull_policy: "Always" - # Docker images will be built and pushed to GCR by default whenever the `scripts/generate_dag.py` is run. To skip building and pushing images, use the optional `--skip-builds` flag. + image: "{{ var.json.chicago_crime.container_registry.run_csv_transform_kub }}" - # Set the environment variables you need initialized in the container. Use these as input variables for the script your container is expected to perform. env_vars: SOURCE_URL: "https://data.cityofchicago.org/api/views/ijzp-q8t2/rows.csv" SOURCE_FILE: "files/data.csv" @@ -79,7 +76,6 @@ dag: TARGET_GCS_BUCKET: "{{ var.json.shared.composer_bucket }}" TARGET_GCS_PATH: "data/chicago_crime/crime/data_output.csv" - # Set resource limits for the pod here. For resource units in Kubernetes, see https://kubernetes.io/docs/concepts/configuration/manage-resources-containers/#resource-units-in-kubernetes resources: request_memory: "8G" request_cpu: "2" @@ -90,25 +86,19 @@ dag: args: task_id: "load_chicago_crime_to_bq" - # The GCS bucket where the CSV file is located in. + bucket: "{{ var.json.shared.composer_bucket }}" - # The GCS object path for the CSV file + source_objects: ["data/chicago_crime/crime/data_output.csv"] source_format: "CSV" destination_project_dataset_table: "chicago_crime.crime" - # Use this if your CSV file contains a header row + skip_leading_rows: 1 - # How to write data to the table: overwrite, append, or write if empty - # See https://cloud.google.com/bigquery/docs/reference/auditlogs/rest/Shared.Types/WriteDisposition write_disposition: "WRITE_TRUNCATE" - # The BigQuery table schema based on the CSV file. For more info, see - # https://cloud.google.com/bigquery/docs/schemas. - # Always use snake_case and lowercase for column names, and be explicit, - # i.e. specify modes for all columns. schema_fields: - name: "unique_key" type: "integer" From f3bf8bd1080d7252f4501ec4919f14832f6e579a Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Fri, 17 Sep 2021 08:36:03 +0000 Subject: [PATCH 4/8] fix: review comments --- .../run_csv_transform_kub/csv_transform.py | 34 ++++++------------- 1 file changed, 10 insertions(+), 24 deletions(-) diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py index ff1463a1c..9df9d9652 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py @@ -34,11 +34,11 @@ def main( ) -> None: logging.info( - "chicago crime process started at " + "Chicago Crime process started at " + str(datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")) ) - logging.info("creating 'files' folder") + logging.info("Creating 'files' folder") pathlib.Path("./files").mkdir(parents=True, exist_ok=True) logging.info(f"Downloading file {source_url}") @@ -58,6 +58,12 @@ def main( logging.info("Transform: Removing null values.. ") filter_null_rows(df) + logging.info("Transform: Converting to integers..") + convert_values_to_integer_string(df) + + logging.info("Transform: Converting to float..") + removing_nan_values(df) + logging.info("Transform: Reordering headers..") df = df[ [ @@ -86,26 +92,6 @@ def main( ] ] - logging.info("Transform: converting to integers..") - - # df["unique_key"] = df["unique_key"].apply(convert_to_integer_string) - # df["beat"] = df["beat"].apply(convert_to_integer_string) - # df["district"] = df["district"].apply(convert_to_integer_string) - # df["ward"] = df["ward"].apply(convert_to_integer_string) - # df["community_area"] = df["community_area"].apply(convert_to_integer_string) - # df["year"] = df["year"].apply(convert_to_integer_string) - - convert_values_to_integer_string(df) - - logging.info("Transform: converting to float..") - - # df["x_coordinate"] = df["x_coordinate"].apply(resolve_nan) - # df["y_coordinate"] = df["y_coordinate"].apply(resolve_nan) - # df["latitude"] = df["latitude"].apply(resolve_nan) - # df["longitude"] = df["longitude"].apply(resolve_nan) - - removing_nan_values(df) - logging.info(f"Saving to output file.. {target_file}") try: save_to_new_file(df, file_path=str(target_file)) @@ -136,7 +122,7 @@ def removing_nan_values(df: pd.DataFrame) -> None: cols = ["x_coordinate", "y_coordinate", "latitude", "longitude"] for cols in cols: - df[cols] = df[cols].apply(convert_dt_format) + df[cols] = df[cols].apply(resolve_nan) def convert_to_integer_string(input: typing.Union[str, float]) -> str: @@ -152,7 +138,7 @@ def convert_values_to_integer_string(df: pd.DataFrame) -> None: cols = ["unique_key", "beat", "district", "ward", "community_area", "year"] for cols in cols: - df[cols] = df[cols].apply(convert_dt_format) + df[cols] = df[cols].apply(convert_to_integer_string) def rename_headers(df: pd.DataFrame) -> None: From ea7d26e8af6b370c983d4000a45a097ef758be15 Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Wed, 29 Sep 2021 14:00:45 +0000 Subject: [PATCH 5/8] fix: Removed white spaces from yaml files --- .../_images/run_csv_transform_kub/Dockerfile | 1 - .../chicago_crime/_terraform/terraform.tfvars | 23 -------------- datasets/chicago_crime/crime/pipeline.yaml | 31 ++----------------- 3 files changed, 2 insertions(+), 53 deletions(-) delete mode 100644 datasets/chicago_crime/_terraform/terraform.tfvars diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile b/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile index 85af90570..7265a1b71 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/Dockerfile @@ -13,7 +13,6 @@ # limitations under the License. # The base image for this build -# FROM gcr.io/google.com/cloudsdktool/cloud-sdk:slim FROM python:3.8 # Allow statements and log messages to appear in Cloud logs diff --git a/datasets/chicago_crime/_terraform/terraform.tfvars b/datasets/chicago_crime/_terraform/terraform.tfvars deleted file mode 100644 index 1ce7806c8..000000000 --- a/datasets/chicago_crime/_terraform/terraform.tfvars +++ /dev/null @@ -1,23 +0,0 @@ -/** - * Copyright 2021 Google LLC - * - * Licensed under the Apache License, Version 2.0 (the "License"); - * you may not use this file except in compliance with the License. - * You may obtain a copy of the License at - * - * http://www.apache.org/licenses/LICENSE-2.0 - * - * Unless required by applicable law or agreed to in writing, software - * distributed under the License is distributed on an "AS IS" BASIS, - * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. - * See the License for the specific language governing permissions and - * limitations under the License. - */ - - -project_id = "bigquery-public-data-dev" -bucket_name_prefix = "public-datasets-dev" -impersonating_acct = "datasets-pipelines-dev-tf@bigquery-public-data-dev.iam.gserviceaccount.com" -region = "us-central1" -env = "dev" - diff --git a/datasets/chicago_crime/crime/pipeline.yaml b/datasets/chicago_crime/crime/pipeline.yaml index 86c382014..812ce8610 100644 --- a/datasets/chicago_crime/crime/pipeline.yaml +++ b/datasets/chicago_crime/crime/pipeline.yaml @@ -17,8 +17,6 @@ resources: - type: bigquery_table table_id: crime - - description: "Chicago Crime dataset" dag: @@ -27,8 +25,6 @@ dag: dag_id: crime default_args: owner: "Google" - - depends_on_past: False start_date: '2021-03-01' max_active_runs: 1 @@ -38,21 +34,12 @@ dag: tasks: - operator: "KubernetesPodOperator" - - description: "Run CSV transform within kubernetes pod" - args: - task_id: "chicago_crime_transform_csv" - startup_timeout_seconds: 600 - - name: "crime" - namespace: "default" - affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: @@ -62,41 +49,27 @@ dag: operator: In values: - "pool-e2-standard-4" - - image_pull_policy: "Always" - - image: "{{ var.json.chicago_crime.container_registry.run_csv_transform_kub }}" - env_vars: SOURCE_URL: "https://data.cityofchicago.org/api/views/ijzp-q8t2/rows.csv" SOURCE_FILE: "files/data.csv" TARGET_FILE: "files/data_output.csv" - TARGET_GCS_BUCKET: "{{ var.json.shared.composer_bucket }}" + TARGET_GCS_BUCKET: "{{ var.value.composer_bucket }}" TARGET_GCS_PATH: "data/chicago_crime/crime/data_output.csv" - resources: request_memory: "8G" request_cpu: "2" - operator: "GoogleCloudStorageToBigQueryOperator" description: "Task to load CSV data to a BigQuery table" - args: task_id: "load_chicago_crime_to_bq" - - - bucket: "{{ var.json.shared.composer_bucket }}" - - + bucket: "{{ var.value.composer_bucket }}" source_objects: ["data/chicago_crime/crime/data_output.csv"] source_format: "CSV" destination_project_dataset_table: "chicago_crime.crime" - - skip_leading_rows: 1 - write_disposition: "WRITE_TRUNCATE" schema_fields: From 177ca5fe0790c07a1dd6a84ae7fd44462f7e437c Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Tue, 5 Oct 2021 13:56:05 +0000 Subject: [PATCH 6/8] feat: changing dataset description --- .../_terraform/chicago_crime_dataset.tf | 2 +- .../_terraform/crime_pipeline.tf | 2 +- datasets/chicago_crime/dataset.yaml | 43 +++++-------------- 3 files changed, 12 insertions(+), 35 deletions(-) diff --git a/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf b/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf index db644febe..9de937909 100644 --- a/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf +++ b/datasets/chicago_crime/_terraform/chicago_crime_dataset.tf @@ -18,7 +18,7 @@ resource "google_bigquery_dataset" "chicago_crime" { dataset_id = "chicago_crime" project = var.project_id - description = "chicago crime" + description = "This dataset reflects reported incidents of crime (with the exception of murders where data exists for each victim) that occurred in the City of Chicago from 2001 to present, minus the most recent seven days. Data is extracted from the Chicago Police Department\u0027s CLEAR (Citizen Law Enforcement Analysis and Reporting) system. In order to protect the privacy of crime victims, addresses are shown at the block level only and specific locations are not identified. This data includes unverified reports supplied to the Police Department. The preliminary crime classifications may be changed at a later date based upon additional investigation and there is always the possibility of mechanical or human error. Therefore, the Chicago Police Department does not guarantee (either expressed or implied) the accuracy, completeness, timeliness, or correct sequencing of the information and the information should not be used for comparison purposes over time.\n\nDataset Source: City of Chicago\n\nCategory: Chicago, Public Safety\n\nUse: This dataset is publicly available for anyone to use under the following terms provided by the Dataset Source \u2014https://data.cityofchicago.org \u2014 and is provided \"AS IS\" without any warranty, express or implied, from Google. Google disclaims all liability for any damages, direct or indirect, resulting from the use of the dataset.\n\nUpdate Frequency: Daily" } output "bigquery_dataset-chicago_crime-dataset_id" { diff --git a/datasets/chicago_crime/_terraform/crime_pipeline.tf b/datasets/chicago_crime/_terraform/crime_pipeline.tf index ab7f09462..a76556314 100644 --- a/datasets/chicago_crime/_terraform/crime_pipeline.tf +++ b/datasets/chicago_crime/_terraform/crime_pipeline.tf @@ -20,7 +20,7 @@ resource "google_bigquery_table" "crime" { dataset_id = "chicago_crime" table_id = "crime" - description = "chicago crime dataset" + description = "Chicago Crime dataset" diff --git a/datasets/chicago_crime/dataset.yaml b/datasets/chicago_crime/dataset.yaml index fc1110d7f..a40007b38 100644 --- a/datasets/chicago_crime/dataset.yaml +++ b/datasets/chicago_crime/dataset.yaml @@ -13,16 +13,8 @@ # limitations under the License. dataset: - # The `dataset` block includes properties for your dataset that will be shown - # to users of your data on the Google Cloud website. - - # Must be exactly the same name as the folder name your dataset.yaml is in. name: chicago_crime - - # A friendly, human-readable name of the dataset friendly_name: chicago_crime - - # A short, descriptive summary of the dataset. description: |- This dataset reflects reported incidents of crime (with the exception of murders where data exists for each victim) that occurred in the City of Chicago from 2001 to present, minus the most recent seven days. Data is extracted from the Chicago Police Department's CLEAR (Citizen Law Enforcement Analysis and Reporting) system. In order to protect the privacy of crime victims, addresses are shown at the block level only and specific locations are not identified. This data includes unverified reports supplied to the Police Department. The preliminary crime classifications may be changed at a later date based upon additional investigation and there is always the possibility of mechanical or human error. Therefore, the Chicago Police Department does not guarantee (either expressed or implied) the accuracy, completeness, timeliness, or correct sequencing of the information and the information should not be used for comparison purposes over time. @@ -33,35 +25,20 @@ dataset: Use: This dataset is publicly available for anyone to use under the following terms provided by the Dataset Source —https://data.cityofchicago.org — and is provided "AS IS" without any warranty, express or implied, from Google. Google disclaims all liability for any damages, direct or indirect, resulting from the use of the dataset. Update Frequency: Daily - - # A list of sources the dataset is derived from, using the YAML list syntax. dataset_sources: ~ - - # A list of terms and conditions that users of the dataset should agree on, - # using the YAML list syntax. terms_of_use: ~ resources: - # A list of Google Cloud resources needed by your dataset. In principle, all - # pipelines under a dataset should be able to share these resources. - # - # The currently supported resources are shown below. Use only the resources - # you need, and delete the rest as needed by your pipeline. - # - # We will keep adding to the list below to support more Google Cloud resources - # over time. If a resource you need isn't supported, please file an issue on - # the repository. - - type: bigquery_dataset - # Google BigQuery dataset to namespace all tables managed by this folder - # - # Required Properties: - # dataset_id - # - # Optional Properties: - # friendly_name (A user-friendly name of the dataset) - # description (A user-friendly description of the dataset) - # location (The geographic location where the dataset should reside) dataset_id: chicago_crime - description: chicago crime + description: |- + This dataset reflects reported incidents of crime (with the exception of murders where data exists for each victim) that occurred in the City of Chicago from 2001 to present, minus the most recent seven days. Data is extracted from the Chicago Police Department's CLEAR (Citizen Law Enforcement Analysis and Reporting) system. In order to protect the privacy of crime victims, addresses are shown at the block level only and specific locations are not identified. This data includes unverified reports supplied to the Police Department. The preliminary crime classifications may be changed at a later date based upon additional investigation and there is always the possibility of mechanical or human error. Therefore, the Chicago Police Department does not guarantee (either expressed or implied) the accuracy, completeness, timeliness, or correct sequencing of the information and the information should not be used for comparison purposes over time. + + Dataset Source: City of Chicago + + Category: Chicago, Public Safety + + Use: This dataset is publicly available for anyone to use under the following terms provided by the Dataset Source —https://data.cityofchicago.org — and is provided "AS IS" without any warranty, express or implied, from Google. Google disclaims all liability for any damages, direct or indirect, resulting from the use of the dataset. + + Update Frequency: Daily From a97b0fc8ed9bf20862b3aded99e6913cb64e4c0f Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Thu, 7 Oct 2021 06:29:31 +0000 Subject: [PATCH 7/8] fix: worked on review comments --- .../run_csv_transform_kub/csv_transform.py | 136 +++++++++++------- datasets/chicago_crime/crime/crime_dag.py | 7 +- datasets/chicago_crime/crime/pipeline.yaml | 5 +- 3 files changed, 89 insertions(+), 59 deletions(-) diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py index 9df9d9652..f95839849 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py @@ -18,6 +18,7 @@ import math import os import pathlib +import subprocess import typing import pandas as pd @@ -31,6 +32,7 @@ def main( target_file: pathlib.Path, target_gcs_bucket: str, target_gcs_path: str, + chunk_size: str, ) -> None: logging.info( @@ -44,59 +46,74 @@ def main( logging.info(f"Downloading file {source_url}") download_file(source_url, source_file) - logging.info(f"Opening file {source_file}") - df = pd.read_csv(source_file) - - logging.info(f"Transforming {source_file} ...") - - logging.info(f"Transform: Rename columns {source_file} ...") - rename_headers(df) - - logging.info("Transform: Converting date format.. ") - convert_values(df) - - logging.info("Transform: Removing null values.. ") - filter_null_rows(df) - - logging.info("Transform: Converting to integers..") - convert_values_to_integer_string(df) - - logging.info("Transform: Converting to float..") - removing_nan_values(df) - - logging.info("Transform: Reordering headers..") - df = df[ - [ - "unique_key", - "case_number", - "date", - "block", - "iucr", - "primary_type", - "description", - "location_description", - "arrest", - "domestic", - "beat", - "district", - "ward", - "community_area", - "fbi_code", - "x_coordinate", - "y_coordinate", - "year", - "updated_on", - "latitude", - "longitude", - "location", - ] - ] - - logging.info(f"Saving to output file.. {target_file}") - try: - save_to_new_file(df, file_path=str(target_file)) - except Exception as e: - logging.error(f"Error saving output file: {e}.") + with pd.read_csv( + source_file, + chunksize=int(chunk_size), + ) as reader: + for chunk_number, chunk in enumerate(reader): + logging.info(f"Processing batch {chunk_number}") + target_file_batch = str(target_file).replace( + ".csv", "-" + str(chunk_number) + ".csv" + ) + df = pd.DataFrame() + df = pd.concat([df, chunk]) + + logging.info(f"Transforming {source_file} ...") + + logging.info(f"Transform: Rename columns {source_file} ...") + rename_headers(df) + + logging.info("Transform: Converting date format.. ") + convert_values(df) + + logging.info("Transform: Removing null values.. ") + filter_null_rows(df) + + logging.info("Transform: Converting to integers..") + convert_values_to_integer_string(df) + + logging.info("Transform: Converting to float..") + removing_nan_values(df) + + logging.info("Transform: Reordering headers..") + df = df[ + [ + "unique_key", + "case_number", + "date", + "block", + "iucr", + "primary_type", + "description", + "location_description", + "arrest", + "domestic", + "beat", + "district", + "ward", + "community_area", + "fbi_code", + "x_coordinate", + "y_coordinate", + "year", + "updated_on", + "latitude", + "longitude", + "location", + ] + ] + + process_chunk(df, target_file_batch) + + logging.info(f"Appending batch {chunk_number} to {target_file}") + if chunk_number == 0: + subprocess.run(["cp", target_file_batch, target_file]) + else: + subprocess.check_call(f"sed -i '1d' {target_file_batch}", shell=True) + subprocess.check_call( + f"cat {target_file_batch} >> {target_file}", shell=True + ) + subprocess.run(["rm", target_file_batch]) logging.info( f"Uploading output file to.. gs://{target_gcs_bucket}/{target_gcs_path}" @@ -109,6 +126,16 @@ def main( ) +def process_chunk(df: pd.DataFrame, target_file_batch: str) -> None: + + logging.info(f"Saving to output file.. {target_file_batch}") + try: + save_to_new_file(df, file_path=str(target_file_batch)) + except Exception as e: + logging.error(f"Error saving output file: {e}.") + logging.info("..Done!") + + def resolve_nan(input: typing.Union[str, float]) -> str: str_val = "" if not input or (math.isnan(input)): @@ -173,7 +200,7 @@ def rename_headers(df: pd.DataFrame) -> None: def convert_dt_format(dt_str: str) -> str: # Old format: MM/dd/yyyy hh:mm:ss aa # New format: yyyy-MM-dd HH:mm:ss - if dt_str is None or len(dt_str) == 0: + if not dt_str: return dt_str else: return datetime.datetime.strptime(dt_str, "%m/%d/%Y %H:%M:%S %p").strftime( @@ -223,4 +250,5 @@ def upload_file_to_gcs(file_path: pathlib.Path, gcs_bucket: str, gcs_path: str) target_file=pathlib.Path(os.environ["TARGET_FILE"]).expanduser(), target_gcs_bucket=os.environ["TARGET_GCS_BUCKET"], target_gcs_path=os.environ["TARGET_GCS_PATH"], + chunk_size=os.environ["CHUNK_SIZE"], ) diff --git a/datasets/chicago_crime/crime/crime_dag.py b/datasets/chicago_crime/crime/crime_dag.py index 8de82c093..97df85d90 100644 --- a/datasets/chicago_crime/crime/crime_dag.py +++ b/datasets/chicago_crime/crime/crime_dag.py @@ -61,16 +61,17 @@ "SOURCE_URL": "https://data.cityofchicago.org/api/views/ijzp-q8t2/rows.csv", "SOURCE_FILE": "files/data.csv", "TARGET_FILE": "files/data_output.csv", - "TARGET_GCS_BUCKET": "{{ var.json.shared.composer_bucket }}", + "TARGET_GCS_BUCKET": "{{ var.value.composer_bucket }}", "TARGET_GCS_PATH": "data/chicago_crime/crime/data_output.csv", + "CHUNK_SIZE": "1000000", }, - resources={"request_memory": "8G", "request_cpu": "2"}, + resources={"request_memory": "2G", "request_cpu": "1"}, ) # Task to load CSV data to a BigQuery table load_chicago_crime_to_bq = gcs_to_bq.GoogleCloudStorageToBigQueryOperator( task_id="load_chicago_crime_to_bq", - bucket="{{ var.json.shared.composer_bucket }}", + bucket="{{ var.value.composer_bucket }}", source_objects=["data/chicago_crime/crime/data_output.csv"], source_format="CSV", destination_project_dataset_table="chicago_crime.crime", diff --git a/datasets/chicago_crime/crime/pipeline.yaml b/datasets/chicago_crime/crime/pipeline.yaml index 812ce8610..de08b4e57 100644 --- a/datasets/chicago_crime/crime/pipeline.yaml +++ b/datasets/chicago_crime/crime/pipeline.yaml @@ -57,9 +57,10 @@ dag: TARGET_FILE: "files/data_output.csv" TARGET_GCS_BUCKET: "{{ var.value.composer_bucket }}" TARGET_GCS_PATH: "data/chicago_crime/crime/data_output.csv" + CHUNK_SIZE: "1000000" resources: - request_memory: "8G" - request_cpu: "2" + request_memory: "2G" + request_cpu: "1" - operator: "GoogleCloudStorageToBigQueryOperator" description: "Task to load CSV data to a BigQuery table" From f6e68b04a14d2c2aa64a5634a2dba20014acb470 Mon Sep 17 00:00:00 2001 From: Dipannita Banerjee Date: Thu, 7 Oct 2021 15:55:17 +0000 Subject: [PATCH 8/8] fix: removing else condition --- .../run_csv_transform_kub/csv_transform.py | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py index f95839849..d7a9f9410 100644 --- a/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py +++ b/datasets/chicago_crime/_images/run_csv_transform_kub/csv_transform.py @@ -137,28 +137,22 @@ def process_chunk(df: pd.DataFrame, target_file_batch: str) -> None: def resolve_nan(input: typing.Union[str, float]) -> str: - str_val = "" if not input or (math.isnan(input)): - str_val = "" - else: - str_val = str(input) - return str_val.replace("None", "") + return "" + return str(input).replace("None", "") def removing_nan_values(df: pd.DataFrame) -> None: cols = ["x_coordinate", "y_coordinate", "latitude", "longitude"] - for cols in cols: df[cols] = df[cols].apply(resolve_nan) def convert_to_integer_string(input: typing.Union[str, float]) -> str: - str_val = "" + if not input or (math.isnan(input)): - str_val = "" - else: - str_val = str(int(round(input, 0))) - return str_val + return "" + return str(int(round(input, 0))) def convert_values_to_integer_string(df: pd.DataFrame) -> None: