Merge branch 'master' into train-test-split-updates

.travis.yml

@@ -1,6 +1,14 @@
+language: python
 dist: xenial
 sudo: required
 
+env:
+  matrix:
+    - PYPI_BUILD="complete"
+    - PYPI_BUILD="only_plotting"
+    - PYPI_BUILD="only_postgres"
+    - PYPI_BUILD="basic"
+
 matrix:
   include:
   - os: linux
@@ -27,8 +35,8 @@ install:
 # os dependencies and set up matplotlib backend
 # travis does not support python on osx - install via miniconda
 - |
-  if [ $TRAVIS_OS_NAME = 'osx' ]; then 
-    echo "Installing brew packages..."  
+  if [ $TRAVIS_OS_NAME = 'osx' ]; then
+    echo "Installing brew packages..."
     brew install graphviz
     echo "Setting up conda environment..."
     wget http://repo.continuum.io/miniconda/Miniconda3-latest-MacOSX-x86_64.sh -O miniconda.sh || exit 1
@@ -51,9 +59,31 @@ install:
   fi
   echo "Installing requirements"
   pip install -r requirements.txt
-  
+  if [ $PYPI_BUILD = 'complete' ]; then
+    echo "Installing postgres and plotting requirements"
+    pip install psycopg2>=2.8.3
+    pip install psycopg2_binary>=2.8.2
+    pip install pygraphviz>=1.5
+  elif [ $PYPI_BUILD = 'only_postgres' ]; then
+    echo "Installing postgres requirements"
+    pip install psycopg2>=2.8.3
+    pip install psycopg2_binary>=2.8.2
+  elif [ $PYPI_BUILD = 'only_plotting' ]; then
+    echo "Installing plotting requirements"
+    pip install pygraphviz>=1.5
+  fi
+
 script:
-- python -m pytest
+  - |
+    if [ $PYPI_BUILD = 'complete' ]; then
+        python -m pytest
+    elif [ $PYPI_BUILD = 'only_plotting' ]; then
+        python -m pytest -m "not postgres"
+    elif [ $PYPI_BUILD = 'only_postgres' ]; then
+        python -m pytest -m "not plotting"
+    else
+        python -m pytest -m "not optional"
+    fi
 
 branches:
   only:

README.md

@@ -12,7 +12,7 @@
 
 ## Primrose at a glance
 
-`Primrose` is a simple **Python** framework for executing **in-memory** workflows defined by directed acyclic graphs (**DAGs**) via configuration files. Data in `primrose` flows from one node to another while **avoiding serialization**, except for when explicitly specified by the user. `Primrose` nodes are designed for **simple batch-based machine learning workflows**, which have datasets small enough to fit into a single machine's memory. 
+`Primrose` is a simple **Python** framework for executing **in-memory** workflows defined by directed acyclic graphs (**DAGs**) via configuration files. Data in `primrose` flows from one node to another while **avoiding serialization**, except for when explicitly specified by the user. `Primrose` nodes are designed for **simple batch-based machine learning workflows**, which have datasets small enough to fit into a single machine's memory.
 
 ## Table of Contents
 We suggest reading the documentation in the following order:
@@ -27,43 +27,49 @@ We suggest reading the documentation in the following order:
 
 ## Introduction
 
- `Primrose` is a Python framework for quick, simple machine learning and recommender deployments developed by the data science team at [WW](https://www.weightwatchers.com/us/). It is essentially a workflow management tool which is specialized for the needs of machine learning tasks with small to medium sized datasets (≤ 100GB). Like many orchestration tools, `Primrose` *nodes* are defined in a [directed-acyclic-graph](https://en.wikipedia.org/wiki/Directed_acyclic_graph) which defines dependencies and control flow. 
- 
+ `Primrose` is a Python framework for quick, simple machine learning and recommender deployments developed by the data science team at [WW](https://www.weightwatchers.com/us/). It is essentially a workflow management tool which is specialized for the needs of machine learning tasks with small to medium sized datasets (≤ 100GB). Like many orchestration tools, `Primrose` *nodes* are defined in a [directed-acyclic-graph](https://en.wikipedia.org/wiki/Directed_acyclic_graph) which defines dependencies and control flow.
+
  Here's an example DAG showing data cleaning, model training, and model serialization:
- 
+
  <p align="center">
    <img src="img/hello_world_tennis.png" width="500">
-</p> 
- 
+</p>
+
  It exists within an ecosystem of other great open source workflow management tools (like [Airflow](https://airflow.apache.org/), [Luigi](https://luigi.readthedocs.io/en/stable/), [Kubeflow](https://www.kubeflow.org/docs/about/kubeflow/) or [Prefect](https://docs.prefect.io/guide/)) while carving it's own niche based on the following design goals:
 
 1. **Avoid unnecessary serialization:** `Primrose` keeps data in-memory between task steps, and only performs (de)serialization operations when explicitly requested by the user. Data is transported between nodes through use of a `DataObject` abstraction, which contextually delivers the correct data to each `Primrose` node at runtime. As a consequence of this design choice, `Primrose` runs on a single machine and can be deployed as a job within a single container, like any other Python script or cron job. In addition to operating on persistent data passed between nodes, `Primrose` can also be used to call external services in a manner similar to a [Luigi](https://luigi.readthedocs.io/en/stable/) job. In this way, Spark jobs or Hadoop scripts can be called and the framework simply dictates dependencies.
 
     * *As a comparison...* many solutions in this space are focused on long-running jobs which may be distributed across several computing nodes. Furthermore, to facilitate parallelization, save states for redundancy, and process datasets which are too large for memory, orchestrators often require that data is serialized between each workflow task. For smaller datasets, the IO time associated with these steps can be much longer than the time spent in computation.
-    
+
     * *Primrose is not...* a solution which scales across clusters or a complex dependency management solution with dynamic DAGs (yet).
 
 2. **Batch processesing for ML:** `Primrose` was built to facilitate frequent batches of model training or predictions that must read and write from/to multiple sources. Rather than requiring users to define their DAG structure in Python code, `Primrose` adopts a `configuration-as-code` approach. `Primrose` users create implementations of node objects once, then any DAG structural modifications or parameterization changes are processed through configuration json files. This way, deployment changes to DAG operations (such as modifying a DAG to serve model predictions instead of training) can be handled purely through configuration files. This avoids the need to build new Python scripts for production modifications. Furthermore, `Primrose` nodes are based on common machine learning tasks to make data scientist's lives easier. This cuts down on development time for building new models and maximizes code re-use among projects and teams. See the modeling examples in the source and documentation for more info!
-    * *As a comparison...* in `Primrose`, users simply need to specify in their configuration file that they want common ML operations to act on the `DataObject`. These ML operations can certainly be implemented by users in Luigi or Airflow, but we found operations such as test-train splits or classifier cross-validation to be so common that they warranted nodes pre-dedicated to these operations.  [Prefect](https://docs.prefect.io/guide/) has made some great strides in this area, and we encourage users to check their solution out. 
-
-    * *Primrose is not...* an auto-ml tool or machine-learning toolkit which implements its own algorithms. Any Python machine learning library can be used with `Primrose`, simply by building model or pipeline nodes that implement the user's choice of library. 
+    * *As a comparison...* in `Primrose`, users simply need to specify in their configuration file that they want common ML operations to act on the `DataObject`. These ML operations can certainly be implemented by users in Luigi or Airflow, but we found operations such as test-train splits or classifier cross-validation to be so common that they warranted nodes pre-dedicated to these operations.  [Prefect](https://docs.prefect.io/guide/) has made some great strides in this area, and we encourage users to check their solution out.
 
-3. **Simplicity:** 
+    * *Primrose is not...* an auto-ml tool or machine-learning toolkit which implements its own algorithms. Any Python machine learning library can be used with `Primrose`, simply by building model or pipeline nodes that implement the user's choice of library.
+
+3. **Simplicity:**
+
+    **Standardization of deployments:** `Primrose` is meant to help make deployment and model building as simple as possible. From a developer operations perspective, it requires no external scheduler or cluster to run deployments. `Primrose` code can simply be containerized with a `primrose` Python entrypoint, and deployed as a job on a k8s or any other container management service.
 
-    **Standardization of deployments:** `Primrose` is meant to help make deployment and model building as simple as possible. From a developer operations perspective, it requires no external scheduler or cluster to run deployments. `Primrose` code can simply be containerized with a `primrose` Python entrypoint, and deployed as a job on a k8s or any other container management service. 
-
     **Standardization of development:** From a software engineering perspective, another advantage of `Primrose` stems form the standardization of model and recommender code. Modifying feature engineering pipelines or adding recommender features is simplified by writing additions to self-contained `Primrose` nodes and making additions to a configuration file.
 
-    * *As a comparison...* `Primrose` can be leveraged as a piece of a larger ETL job (a `Primrose` job could be a job within an Airflow DAG), or run on it's own as a self-contained, single node ETL job. Some orchestration solutions (Airflow, for example) require running persistent clusters and services for managing jobs. 
-     
-    * *Primrose is not...* able to manage its own job scheduling or timing. This is left to user using k8s job scheduling or manual cron job assignments on a virtual machine.  
+    * *As a comparison...* `Primrose` can be leveraged as a piece of a larger ETL job (a `Primrose` job could be a job within an Airflow DAG), or run on it's own as a self-contained, single node ETL job. Some orchestration solutions (Airflow, for example) require running persistent clusters and services for managing jobs.
+
+    * *Primrose is not...* able to manage its own job scheduling or timing. This is left to user using k8s job scheduling or manual cron job assignments on a virtual machine.
 
 
 There are many solutions in this space, and we encourage users to explore other options that may be most appropriate for their workflows. We view `Primrose` as a simple solution for managing production ML jobs.
 
 
 ## Getting Started
 
+`Primrose` has a couple of optional tools:
+* a postgres database reader
+* a plotting tool
+
+These require a few external dependencies, prior to its installation. If interested in their functionality, follow the appropriate instructions for your OS below. Otherwise, you can proceed with the basic package installation.
+
 ### Installation
 
 You can install the latest `Primrose` release via pypi
@@ -77,7 +83,51 @@ cd primrose
 python setup.py install
 ```
 
-`Primrose` has a few external dependencies which need to be installed in addition to the python requirements. Follow the appropriate instructions for your OS below.
+To install the complete `Primrose` package (after dependencies have been installed):
+```
+    pip install primrose[postgres, plotting]
+```
+
+To install `Primrose` with just the postgres option:
+
+```
+pip install primrose[postgres]
+```
+
+To install `primrose` with just the plotting option:
+
+```
+pip install primrose[plotting]
+```
+
+### External dependenices
+
+**Postgres**
+
+#### MacOSX
+
+We recommend using homebrew to manage OS level external packages. If you do not already have homebrew installed, please visit [their website](https://brew.sh/).
+
+Instructions:
+1. Use homebrew to install `postgresql` library.
+   ```
+   brew install postgresql
+   ```
+
+2. Use `pip` to install `psycopg2`
+   ```
+    pip install psycopg2
+   ```
+
+#### Debian/Ubuntu
+
+Instructions:
+1. Install the `libpq-dev` library
+   ```
+   apt-get install libpq-dev
+   ```
+
+**Plotting**
 
 #### MacOSX
 

primrose/configuration/configuration_dag.py

@@ -13,6 +13,7 @@
 from primrose.node_factory import NodeFactory
 from primrose.base.conditional_path_node import AbstractConditionalPath
 
+
 class ConfigurationDag():
 
     def __init__(self, config):
@@ -113,7 +114,7 @@ def descendents(self, source):
 
     def paths(self, source, target):
         """return the paths, if any, from a given source node to a given target node
-            
+
         Args:
             source (str): name of node which is starting point of path
             target (str): name of node which is end point of path
@@ -156,7 +157,7 @@ def check_connected_components(self):
             ConfigurationError if multiuple connected components
 
         """
-        connected_components = nx.connected_components(self.G) 
+        connected_components = nx.connected_components(self.G)
         n = sum([1 for c in connected_components])
         if n > 1:
             raise ConfigurationError("Found multiple connected components: %s" % str(list( nx.connected_components(self.G) )))
@@ -279,7 +280,7 @@ def check_dag(self):
 
         self.check_connected_components()
 
-        self.check_for_cycles() 
+        self.check_for_cycles()
 
     def plot_dag(self, filename, traverser, node_size=500, label_font_size=12, text_angle=0, image_width=16, image_height=12):
         """plot the DAG to image file
@@ -289,7 +290,7 @@ def plot_dag(self, filename, traverser, node_size=500, label_font_size=12, text_
             title (str): title to add to chart
             node_size (int): node size
             label_font_size (int): font size
-            text_angle (int): angle to rotate. This is angle in degrees counter clockwise from east 
+            text_angle (int): angle to rotate. This is angle in degrees counter clockwise from east
             image_width (int): width of image in inches
             image_height (int): heightof image in inches
 
@@ -321,7 +322,10 @@ def plot_dag(self, filename, traverser, node_size=500, label_font_size=12, text_
             import pydot
             from networkx.drawing.nx_pydot import graphviz_layout
         except ImportError: # pragma: no cover
-            raise ImportError("This example needs Graphviz and pydot")
+            raise ImportError(
+                "This example needs Graphviz and pydot."
+                "Please refer to the Plotting requirements in the README"
+            )
 
         # pos = nx.spring_layout(G)
         # pos = nx.circular_layout(G)

primrose/readers/postgres_helper.py

@@ -1,15 +1,22 @@
 """Postgres helper
 
-Author(s): 
+Author(s):
     Wassym Kalouache ([email protected])
 
     Carl Anderson ([email protected])
 
 """
 import os
-import psycopg2
 from primrose.readers.database_helper import get_env_val
 
+try:
+    import psycopg2
+    HAS_PSYCOPG2 = True
+
+except ImportError:
+    HAS_PSYCOPG2 = False
+
+
 class PostgresHelper():
     '''
         some utility methods for connecting to postgres
@@ -43,6 +50,9 @@ def create_db_connection():
                 db (connection): postgres db object
 
         """
+        if not HAS_PSYCOPG2:
+            raise ImportError("psycopg2 is necessary to establish connection")
+
         host, port, username, password, database = PostgresHelper.extract_postgres_credentials()
         conn = psycopg2.connect(dbname=database, user=username, password=password, host=host, port=port, sslmode='require')
         return conn

requirements.txt

@@ -10,15 +10,12 @@ pytest~=3.10.1
 testfixtures>=6.8.2
 boto3>=1.9.0
 matplotlib>=2.1.2
-psycopg2_binary>=2.8.2
 protobuf>=3.8.0
 mysql-connector-python>=8.0.16
-psycopg2>=2.8.3
 pytest-cov>=2.6.0
 scikit-learn>=0.21.2
 atlas>=0.27.0
 google-cloud-storage>=1.16.0
-pygraphviz>=1.5
 jstyleson==0.0.2
 Jinja2>=2.10
 PyYAML>=5.1

setup.py

@@ -35,4 +35,8 @@
         'Source': 'https://github.com/ww-tech/primrose',
       },
       entry_points={"console_scripts": ["primrose = primrose.__init__:cli"]},
+      extras_require={
+        'postgres': ["psycopg2>=2.8.3", "psycopg2_binary>=2.8.2"],
+        'plotting': ["pygraphviz>=1.5"]
+      }
     ),

test/test_configuration_dag.py

@@ -151,7 +151,8 @@ def test_type_bad2():
         dag.create_dag()
     assert "Did not find doesnotexist destination in" in str(e)
 
-
+@pytest.mark.optional
+@pytest.mark.plotting
 def test_plot_dag():
 
     class TestPostprocess(AbstractNode):
@@ -348,7 +349,7 @@ def test_starting_nodes():
                     "filename": "hello_world_predictions.csv"
                 }
             }
-        }   
+        }
     }
     configuration = Configuration(config_location=None, is_dict_config=True, dict_config=config)
 
@@ -394,7 +395,7 @@ def test_nodes_of_type():
                     "filename": "hello_world_predictions.csv"
                 }
             }
-        }   
+        }
     }
     configuration = Configuration(config_location=None, is_dict_config=True, dict_config=config)
 
@@ -447,7 +448,7 @@ def test_paths():
                     "filename": "hello_world_predictions.csv"
                 }
             }
-        }   
+        }
     }
     configuration = Configuration(config_location=None, is_dict_config=True, dict_config=config)
 

test/test_postgres_reader.py

@@ -11,6 +11,9 @@
 def test_necessary_config():
     assert len(PostgresReader.necessary_config({})) == 1
 
+
+@pytest.mark.optional
+@pytest.mark.postgres
 def test_run(monkeypatch):
     config = {
         "implementation_config": {
@@ -52,4 +55,4 @@ def fake_df(query, con):
         dd = data_object.get('mynode', rtype=DataObjectResponseType.KEY_VALUE.value)
         assert 'query_0' in dd
         df = dd['query_0']
-        assert list(df.T.to_dict().values())[0] == {'Name': "Tom", "Age": 20}
+        assert list(df.T.to_dict().values())[0] == {'Name': "Tom", "Age": 20}