updating README.md

.gitignore

@@ -0,0 +1,145 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# USER
+
+# ignore the data csv files for input/output
+**/data/**/*.csv
+**/data/**/*.json
+.vscode/settings.json

.python-version

@@ -0,0 +1 @@
+3.7.9

README.md

@@ -1,2 +1,9 @@
 # street_group_tech_test
-Technical Test for Street Group
+
+Technical Test for Street Group for Daniel Tomlinson.
+
+## Documentation
+
+Read the documentation on github pages for instructions around running the code and a discussion on the approach.
+
+https://dtomlinson91.github.io/street_group_tech_test/

analyse_properties/__init__.py

@@ -0,0 +1,5 @@
+"""
+Daniel Tomlinson (dtomlinson@panaetius.co.uk).
+
+Technical test for Street Group.
+"""

analyse_properties/debug.py

@@ -0,0 +1,24 @@
+import apache_beam as beam
+
+
+class DebugShowEmptyColumn(beam.DoFn):
+    def __init__(self, index):
+        self.index = index
+
+    def process(self, element):
+        column = element[self.index]
+        if len(column) == 0:
+            yield element
+        return None
+
+
+class DebugShowColumnWithValueIn(beam.DoFn):
+    def __init__(self, index, value):
+        self.index = index
+        self.value = value
+
+    def process(self, element):
+        column = element[self.index]
+        if self.value in column:
+            yield element
+        return None

analyse_properties/main.py

@@ -0,0 +1,285 @@
+from datetime import datetime
+import hashlib
+import itertools
+import pathlib
+
+import apache_beam as beam
+
+# from analyse_properties.debug import DebugShowEmptyColumn, DebugShowColumnWithValueIn
+
+
+def slice_by_range(element, *ranges):
+    """Slice a list with multiple ranges."""
+    return itertools.chain(*(itertools.islice(element, *r) for r in ranges))
+
+
+class DropRecordsSingleEmptyColumn(beam.DoFn):
+    """If a given item in a list is empty, drop this entry from the PCollection."""
+
+    def __init__(self, index):
+        self.index = index
+
+    def process(self, element):
+        column = element[self.index]
+        if len(column) == 0:
+            return None
+        yield element
+
+
+class DropRecordsTwoEmptyColumn(beam.DoFn):
+    """If two given items in a list are both empty, drop this entry from the PCollection."""
+
+    def __init__(self, index_0, index_1):
+        self.index_0 = index_0
+        self.index_1 = index_1
+
+    def process(self, element):
+        column_0 = element[self.index_0]
+        column_1 = element[self.index_1]
+        if len(column_0) == 0 and len(column_1) == 0:
+            return None
+        yield element
+
+
+class SplitColumn(beam.DoFn):
+    """Split an item in a list into two separate items in the PCollection."""
+
+    def __init__(self, index, split_char):
+        self.index = index
+        self.split_char = split_char
+
+    def process(self, element):
+        # If there is a split based on the split_char, then keep the first result in
+        # place and append the second.
+        try:
+            part_0, part_1 = element[self.index].split(self.split_char)
+            element[self.index] = part_1.strip()
+            element.append(part_0.strip())
+            yield element
+        except ValueError:
+            element.append("")
+            yield element
+
+
+class GenerateUniqueID(beam.DoFn):
+    """
+    Generate a unique ID for the PCollection, either for all the columns or for the
+    uniquely identifying data only.
+    """
+
+    def __init__(self, all_columns=False):
+        self.all_columns = all_columns
+
+    def process(self, element):
+        unique_string = (
+            ",".join(element[2:]) if not self.all_columns else ",".join(element)
+        )
+        hashed_string = hashlib.md5(unique_string.encode())
+        # append the hash to the end
+        element.append(hashed_string.hexdigest())
+        yield element
+
+
+class DeduplicateByID(beam.DoFn):
+    """
+    If the PCollection has multiple entries after being grouped by ID for all columns,
+    deduplicate the list to keep only one.
+    """
+
+    def process(self, element):
+        if len(element[1]) > 0:
+            deduplicated_element = (element[0], [element[1][0]])
+            yield deduplicated_element
+        else:
+            yield element
+
+
+class RemoveUniqueID(beam.DoFn):
+    """Remove the unique ID from the PCollection, transforming it back into a list."""
+
+    def process(self, element):
+        element_no_id = element[-1][0]
+        element_no_id.pop(-1)
+        yield element_no_id
+
+
+class ConvertDataToDict(beam.DoFn):
+    """Convert the processed data into a dict to be exported as a JSON object."""
+
+    @staticmethod
+    def get_latest_transaction(transaction_dates):
+        """Get the date of the latest transaction."""
+        transaction_dates = [
+            datetime.strptime(individual_transaction, "%Y-%m-%d")
+            for individual_transaction in transaction_dates
+        ]
+        return max(transaction_dates).strftime("%Y")
+
+    @staticmethod
+    def get_readable_address(address_components: list, address_comparisons: list):
+        # Get pairwise comparison to see if two locality/town/district/counties
+        # are equivalent
+        pairwise_comparison = [
+            x == y
+            for i, x in enumerate(address_comparisons)
+            for j, y in enumerate(address_comparisons)
+            if i > j
+        ]
+        # Create a mask to eliminate the redundant parts of the address
+        mask = [True, True, True, True]
+        if pairwise_comparison[0]:
+            mask[1] = False
+        if pairwise_comparison[1] or pairwise_comparison[2]:
+            mask[2] = False
+        if pairwise_comparison[3] or pairwise_comparison[4] or pairwise_comparison[5]:
+            mask[3] = False
+        # Apply the mask
+        applied_mask = list(itertools.compress(address_comparisons, mask))
+        # Filter out empty items in list
+        deduplicated_address_part = list(filter(None, applied_mask))
+
+        # Filter out any missing parts of the address components
+        cleaned_address_components = list(filter(None, address_components))
+
+        # Return the readable address
+        return "\n".join(
+            itertools.chain.from_iterable(
+                [
+                    cleaned_address_components[0:-1],
+                    deduplicated_address_part,
+                    [cleaned_address_components[-1]],
+                ]
+            )
+        )
+
+    def process(self, element):
+        # Group together all the transactions for the property.
+        property_transactions = [
+            {
+                "price": entry[0],
+                "transaction_date": entry[1].replace(" 00:00", ""),
+                "year": entry[1][0:4],
+            }
+            for entry in element[-1]
+        ]
+
+        # Create the dict to hold all the information about the property.
+        json_object = {
+            "property_id": element[0],
+            "readable_address": None,
+            "flat_appartment": element[-1][0][4],
+            "builing": element[-1][0][10],
+            "number": element[-1][0][3],
+            "street": element[-1][0][5],
+            "locality": element[-1][0][6],
+            "town": element[-1][0][7],
+            "district": element[-1][0][8],
+            "county": element[-1][0][9],
+            "postcode": element[-1][0][2],
+            "property_transactions": property_transactions,
+            "latest_transaction_year": self.get_latest_transaction(
+                [
+                    transaction["transaction_date"]
+                    for transaction in property_transactions
+                ]
+            ),
+        }
+
+        # Create a human readable address to go in the dict.
+        json_object["readable_address"] = self.get_readable_address(
+            [
+                json_object["flat_appartment"],
+                json_object["builing"],
+                f'{json_object["number"]} {json_object["street"]}',
+                json_object["postcode"],
+            ],
+            [
+                json_object["locality"],
+                json_object["town"],
+                json_object["district"],
+                json_object["county"],
+            ],
+        )
+        yield json_object
+
+
+def main():
+    # Load in the data from a csv file.
+    input_file = (
+        pathlib.Path(__file__).parents[1]
+        / "data"
+        / "input"
+        / "pp-monthly-update-new-version.csv"
+    )
+
+    with beam.Pipeline() as pipeline:
+        # Load the data
+        load = (
+            pipeline
+            | "Read input data" >> beam.io.ReadFromText(str(input_file))
+            | "Split by ','" >> beam.Map(lambda element: element.split(","))
+            | "Remove leading and trailing quotes"
+            >> beam.Map(lambda element: [el.strip('"') for el in element])
+        )
+
+        # Clean the data by dropping unneeded rows.
+        clean_drop = (
+            load
+            | "Drop unneeded columns"
+            >> beam.Map(lambda element: list(slice_by_range(element, (1, 4), (7, 14))))
+            | "Convert to Upper Case"
+            >> beam.Map(lambda element: [e.upper() for e in element])
+            | "Strip leading/trailing whitespace"
+            >> beam.Map(lambda element: [e.strip() for e in element])
+            | "Drop Empty Postcodes" >> beam.ParDo(DropRecordsSingleEmptyColumn(2))
+            | "Drop empty PAON if missing SAON"
+            >> beam.ParDo(DropRecordsTwoEmptyColumn(3, 4))
+            | "Split PAON into two columns if separated by comma"
+            >> beam.ParDo(SplitColumn(3, ","))
+        )
+
+        # Clean the data by creating an ID, and deduplicating to eliminate repeated rows.
+        clean_deduplicate = (
+            clean_drop
+            | "Generate unique ID for all columns"
+            >> beam.ParDo(GenerateUniqueID(all_columns=True))
+            | "Group by the ID for all columns"
+            >> beam.GroupBy(lambda element: element[-1])
+            | "Deduplicate by the ID for all columns" >> beam.ParDo(DeduplicateByID())
+        )
+
+        # Prepare the data by generating an ID using the uniquely identifying information only
+        # and grouping them by this ID.
+        prepare = (
+            clean_deduplicate
+            | "Remove previous unique ID" >> beam.ParDo(RemoveUniqueID())
+            | "Generate unique ID ignoring price & date"
+            >> beam.ParDo(GenerateUniqueID())
+            | "Group by the ID ignoring price & date"
+            >> beam.GroupBy(lambda element: element[-1])
+        )
+
+        # Format the data into a dict.
+        formatted = (
+            prepare
+            | "Convert the prepared data into a dict object"
+            >> beam.ParDo(ConvertDataToDict())
+        )
+
+        # Save the data to a .json file.
+        output_file = (
+            pathlib.Path(__file__).parents[1] / "data" / "output" / "pp-complete"
+        )
+        output = (
+            formatted
+            | "Combine into one PCollection" >> beam.combiners.ToList()
+            | "Save to .json file"
+            >> beam.io.WriteToText(
+                file_path_prefix=str(output_file),
+                file_name_suffix=".json",
+            )
+        )
+
+
+if __name__ == "__main__":
+    main()

docs/index.md

@@ -0,0 +1,5 @@
+# Welcome
+
+## Introduction
+
+This documentation accompanies the technical test for the Street Group.

download_data.sh

@@ -0,0 +1,5 @@
+# Full data set
+# wget https://storage.googleapis.com/street-group-technical-test-dmot/pp-complete.csv -P data/input
+
+# Monthly update data set
+wget https://storage.googleapis.com/street-group-technical-test-dmot/pp-monthly-update-new-version.csv -P data/input

exploration/report.py

@@ -0,0 +1,35 @@
+from importlib import resources
+
+import pandas as pd
+from pandas_profiling import ProfileReport
+
+
+def main():
+    with resources.path("analyse_properties.data", "pp-complete.csv") as csv_file:
+        df_report = pd.read_csv(
+            csv_file,
+            names=[
+                "transaction_id",
+                "price",
+                "date_of_transfer",
+                "postcode",
+                "property_type",
+                "old_new",
+                "duration",
+                "paon",
+                "saon",
+                "street",
+                "locality",
+                "town_city",
+                "district",
+                "county",
+                "ppd_category",
+                "record_status",
+            ],
+        )
+    profile = ProfileReport(df_report, title="Price Paid Data", minimal=True)
+    profile.to_file("price_paid_data_report.html")
+
+
+if __name__ == "__main__":
+    main()

mkdocs.yaml

@@ -0,0 +1,40 @@
+site_name: The Street Group Technical Test
+repo_url: https://github.com/dtomlinson91/street_group_tech_test
+use_directory_urls: false
+nav:
+  - Documentation:
+      - Welcome: index.md
+  #     - Installation: documentation/installation.md
+  #     - Usage: documentation/usage.md
+  # - Comments and Caveats:
+  #     - Introduction: comments_caveats/introduction.md
+  #     - Time limit: comments_caveats/time_limit.md
+  #     - Third party libraries: comments_caveats/third_party_libraries.md
+  #     - Areas of improvement and comments: comments_caveats/area_of_improvement_comments.md
+  #     - Similar names algorithm: comments_caveats/similar_names.md
+  # - Reference:
+  #     - deduplicator.main: reference/api_documentation_main.md
+  # - Changelog: changelog/changelog.md
+theme:
+  name: material
+  palette:
+    primary: indigo
+    accent: blue
+  feature:
+    tabs: true
+markdown_extensions:
+  - admonition
+  - codehilite:
+      guess_lang: true
+  - toc:
+      permalink: true
+  - pymdownx.superfences
+  # - pymdownx.arithmatex:
+  #     generic: true
+plugins:
+  - search:
+      lang: en
+extra_javascript:
+  - javascripts/config.js
+  - https://polyfill.io/v3/polyfill.min.js?features=es6
+  - https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js

notes/documentation/answers.md

@@ -0,0 +1,7 @@
+# Answers
+
+## CSV
+
+Read a csv file into Beam:
+
+<https://stackoverflow.com/a/41171867>

notes/documentation/beam_docs.md

@@ -0,0 +1,12 @@
+# Beam Documentation
+
+## Transforms
+
+FlatMap:
+
+<https://beam.apache.org/documentation/transforms/python/elementwise/flatmap/>
+
+
+I/O Transforms:
+
+<https://beam.apache.org/documentation/io/built-in/>

notes/links.md

@@ -0,0 +1,5 @@
+# Links
+
+## Data
+
+https://www.gov.uk/government/statistical-data-sets/price-paid-data-downloads

prospector.yaml

@@ -0,0 +1,78 @@
+# This will enable almost every single warning
+output-format: vscode
+doc-warnings: true
+# allow-shorthand: false
+strictness: none
+
+ignore-patterns:
+  - (^|/)\..+
+
+pylint:
+  run: true
+  disable:
+    - fixme
+    - bad-continuation
+    - missing-module-docstring
+    - logging-fstring-interpolation
+    - missing-function-docstring
+    - abstract-method
+    - missing-class-docstring
+    - super-init-not-called
+    - arguments-differ
+    - inconsistent-return-statements
+  enable:
+
+  options:
+    max-locals: 15
+    max-returns: 6
+    max-branches: 12
+    max-statements: 50
+    # max-parents: 7
+    max-attributes: 20
+    min-public-methods: 0
+    max-public-methods: 25
+    max-module-lines: 1000
+    max-line-length: 88
+
+mccabe:
+  run: true
+  options:
+    max-complexity: 10
+
+pep8:
+  run: true
+  options:
+    max-line-length: 88
+    single-line-if-stmt: n
+  disable:
+    - E501 # line too long
+
+pyroma:
+  run: false
+  disable:
+    - PYR19
+    - PYR16
+
+pep257:
+  disable:
+    - D000
+    - D203
+    # - D213
+    - D212 # multiline docstrings on first line only.
+    - D404
+    - D100
+    - D407 # numpy docstring format
+    - D107 # missing docstring in __init__
+    # Docstrings ending with newlines and : in Returns block.
+    - D413
+    - D406
+    - D103
+    - D101 # missing docstring in public class
+    - D102 # missing docstring in public method
+
+pyflakes:
+  disable:
+    - F401 # unused import
+
+dodgy:
+  run: true

pyproject.toml

@@ -0,0 +1,20 @@
+[tool.poetry]
+name = "analyse_properties"
+version = "0.1.0"
+description = ""
+authors = ["Daniel Tomlinson <dtomlinson@panaetius.co.uk>"]
+
+[tool.poetry.dependencies]
+python = "^3.7"
+apache-beam = {extras = ["gcp"], version = "^2.32.0"}
+mkdocs = "^1.2.2"
+mkdocs-material = "^7.3.0"
+
+[tool.poetry.dev-dependencies]
+# pytest = "^5.2"
+prospector = "^1.5.1"
+pandas-profiling = "^3.0.0"
+
+[build-system]
+requires = ["poetry-core>=1.0.0"]
+build-backend = "poetry.core.masonry.api"