Python

AutoML – using TPOT

Posted on Updated on

Another popular AutoML library is TPOT, which stands for Tree-Based Pipeline Optimization Tool. The goal of TPOT is to automate the building of ML pipelines by combining a flexible expression tree representation of pipelines with stochastic search algorithms such as genetic programming. TPOT makes use of the Python-based scikit-learn library

Install the TPOT library using

pip3 install tpot

Here is an example tree-based pipeline from TPOT. Each circle corresponds to a machine learning operator, and the arrows indicate the direction of the data flow

Fig. 8.1

Let’s build upon my previous blog post on AutomML, by using the same data set, with no modifications, and using the training (X_train, y_train) and test (X_test, y_test) data sets (dataframes), based on the Bank data sets. Check the previous post for the detailed steps on getting to this point.

In a similar way as the autosklean library example, I’m just going to demonstrate using TPOT for a classification problem using TPOTClassifier class. For regression problems, there is the corresponding TPOTRegressor class (not demonstrated in this post).

TPOTClassifier has the following main parameters (there are others):

  • generations: Number of iterations to the run pipeline optimization process. The default is 100.
  • population_size: Number of individuals to retain in the genetic programming population every generation. The default is 100.
  • offspring_size: Number of offspring to produce in each genetic programming generation. The default is 100.
  • mutation_rate: Mutation rate for the genetic programming algorithm in the range [0.0, 1.0]. This parameter tells the GP algorithm how many pipelines to apply random changes to every generation. Default is 0.9
  • crossover_rate: Crossover rate for the genetic programming algorithm in the range [0.0, 1.0]. This parameter tells the genetic programming algorithm how many pipelines to “breed” every generation.
  • scoring: Function used to evaluate the quality of a given pipeline for the classification problem like accuracy, average_precision, roc_auc, recall, etc. The default is accuracy.
  • cv: Cross-validation strategy used when evaluating pipelines. The default is 5.
  • random_state: The seed of the pseudo-random number generator used in TPOT. Use this parameter to make sure that TPOT will give you the same results each time you run it against the same data set with that seed.
  • verbosity: How much information TPOT communicates while it is running. Default is 0 (zero) TPOT will display nothing. 1=display minimal information, 2=display more information and progress bar, 3=print everything and progress bar.
  • n_jobs: Number of processes to use. Default is 1. Use -1 to use all available cores.

Care is needed with some of these settings, for example generations should be set small to begin with, for example set to 5 initially. Also, population_size should also be kept small, for example 5 initially. These initial settings will evaluate 25 piplelines (5×5) configurations before finishing, and for some these settings may need to be adjusted smaller for initial work/investigations. Another parameter to adjust is the ‘verbosity’ setting. The default is 0 which means no details will be displayed. I like to set this to 3, as it gives more details of the outcomes from each pipeline. Adjust higher for more details or lower to fewer details. Another parameter to consider adjusting is ‘max_time_min’ and ‘max_eval_time_min’, but setting these too low can result in no or minimum results.

Load the library, setup the configuration and run. This is very simple to setup

from tpot import TPOTClassifier

#configure settings
tpot = TPOTClassifier(generations=5, population_size=5, verbosity=3, n_jobs=4, scoring='accuracy')

#run TPOT
tpot.fit(X_train, y_train)

As verbosity is set to 3 we get a lot of detail being displayed for each generation. The final output is shown below. What is missing from this is the progress bars which are displayed while TPOT is running

32 operators have been imported by TPOT.


Generation 1 - Current Pareto front scores:

-1	0.8963961891371728	RandomForestClassifier(input_matrix, RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=5, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

-2	0.8978183008194085	RandomForestClassifier(ZeroCount(input_matrix), RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=5, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)
Pipeline encountered that has previously been evaluated during the optimization process. Using the score from the previous evaluation.

Generation 2 - Current Pareto front scores:

-1	0.8974020496851336	RandomForestClassifier(input_matrix, RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=8, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

-2	0.8978183008194085	RandomForestClassifier(ZeroCount(input_matrix), RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=5, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)
_pre_test decorator: _random_mutation_operator: num_test=0 '(slice(None, None, None), 0)' is an invalid key.
Pipeline encountered that has previously been evaluated during the optimization process. Using the score from the previous evaluation.

Generation 3 - Current Pareto front scores:

-1	0.8974020496851336	RandomForestClassifier(input_matrix, RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=8, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

-2	0.8978183008194085	RandomForestClassifier(ZeroCount(input_matrix), RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=5, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)
Skipped pipeline #21 due to time out. Continuing to the next pipeline.
Skipped pipeline #23 due to time out. Continuing to the next pipeline.

Generation 4 - Current Pareto front scores:

-1	0.8974020496851336	RandomForestClassifier(input_matrix, RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=8, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

-2	0.8978183008194085	RandomForestClassifier(ZeroCount(input_matrix), RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.7000000000000001, RandomForestClassifier__min_samples_leaf=5, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

Generation 5 - Current Pareto front scores:

-1	0.8983385200075953	RandomForestClassifier(input_matrix, RandomForestClassifier__bootstrap=True, RandomForestClassifier__criterion=gini, RandomForestClassifier__max_features=0.55, RandomForestClassifier__min_samples_leaf=8, RandomForestClassifier__min_samples_split=7, RandomForestClassifier__n_estimators=100)

TPOTClassifier(generations=5, n_jobs=4, population_size=5, scoring='accuracy',
               verbosity=3)

We can now display the ‘best’ model configuration discovered by TPOT.

tpot.fitted_pipeline_


Pipeline(steps=[('normalizer', Normalizer(norm='l1')),
                ('xgbclassifier',
                 XGBClassifier(base_score=0.5, booster='gbtree',
                               colsample_bylevel=1, colsample_bynode=1,
                               colsample_bytree=1, gamma=0, gpu_id=-1,
                               importance_type='gain',
                               interaction_constraints='', learning_rate=0.01,
                               max_delta_step=0, max_depth=8,
                               min_child_weight=7, missing=nan,
                               monotone_constraints='()', n_estimators=100,
                               n_jobs=1, num_parallel_tree=1, random_state=0,
                               reg_alpha=0, reg_lambda=1, scale_pos_weight=1,
                               subsample=0.8, tree_method='exact',
                               validate_parameters=1, verbosity=0))])

In this run of TPOT, on this data set, XGBoost algorithm gave the best results using the parameters and settings listed above. What is interesting, everytime I’ve run TPOT for the same data set, using the same configuration parameters, I get a slightly different outcome.

Next step is to evaluate the ‘best’ model on the holdout data set.

tpot.score(X_test, y_test)


0.9037792344420167

The results achieved are good and are better than some of the other models created by other AutoML libraries.

The final step we can perform is to export the model template. This creates a file containing the template code to create and use the model. This does require some modifications to specify the data set, and the pipeline of data modifications and transformations.

#export the model
tpot.export('.../tpot_Bank_pipeline.py')

The output file contains the following.

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.pipeline import make_pipeline
from sklearn.preprocessing import Normalizer
from xgboost import XGBClassifier

# NOTE: Make sure that the outcome column is labeled 'target' in the data file
tpot_data = pd.read_csv('PATH/TO/DATA/FILE', sep='COLUMN_SEPARATOR', dtype=np.float64)
features = tpot_data.drop('target', axis=1)

training_features, testing_features, training_target, testing_target = \
            train_test_split(features, tpot_data['target'], random_state=None)

# Average CV score on the training set was: 0.8986507248984001
exported_pipeline = make_pipeline(
    Normalizer(norm="l1"),
    XGBClassifier(learning_rate=0.01, max_depth=8, min_child_weight=7, n_estimators=100, n_jobs=1, subsample=0.8, verbosity=0)
)

exported_pipeline.fit(training_features, training_target)
results = exported_pipeline.predict(testing_features)

TPOT does have some issues and limitations. Well it is slow, and part of this is due to the nature of genetic algorithms, every time you run TPOT you may get different results, etc. Some of these issues can be addressed by adjusting some of the parameters, but even still, it doesn’t eliminate all of them. Running on GPU helps a little with timing of each run. TPOT doesn’t remove the need for data cleaning, feature engineering etc, but that is the case with most solutions.

AutoML – using autosklearn in Python

Posted on Updated on

I’ve written some previous posts about AutoML and how to use AutoML with Oracle OML4Py (part 1 and part 2) and AutoML UI.

Building upon these, in this post I’ll demonstrate how to use autosklearn Python Package to do something similar, using the same data set I used in my previous posts.

To install the package run the typical pip command

pip3 install auto-sklearn

I did have some challegenges with installing this package, and this seems to be common, with different people having slightly different issues. These mainly revolved around having to install/update the swiff and pyrfr Python packages. Once done, then autosklearn package installed.

Let’s do a simple test

import autosklearn
print('autosklearn: %s' % autosklearn.__version__)

autosklearn: 0.12.5

Just like in my previous examples, I’m just going to use autosklearn to build a Classification model, as that is what the data set is designed for.

from sklearn.metrics import accuracy_score

# define search
model = autosklearn.classification.AutoSklearnClassifier()
# perform the search
model.fit(X_train, y_train)

The code above is a very basic configuration, and if this is the first time you are going to run this, then DON’T. There are a lot of parameter you can set, with one of them being ‘time_left_for_this_task’. The default value for this parameter is 360, which is one hour. Not a good idea! Set this to being much lower, say for an initial run of 3-5 minutes. This should be enough time for it to build many different models. I like to set the time for this using a multiplier of 60 (seconds). That way you don’t have to do any calculations! Two other parameters to consider setting/changing are

  • n_jobs: this is the number of jobs to run in parallel. Default is -1, which uses all processors, or set to to a number, eg. 4
  • metric: what evaluation metric to use for the models. For classification we have, accuracy, balanced_accuracy, f1, f1_marco, f1_micro, f1_samples, f1_weighted, roc_auc, precision, precision_macro, precision_micro, precision_samples, precision_weighted, average_percision, recall, recall_macro, recall_micro, recall_samples, recall_weighted and log_loss. For regression problems, r2, mean_squared_error, mean_absolute_error and median_absolute_error

Using these parameters let’s run a search.

# define search
model2 = autosklearn.classification.AutoSklearnClassifier(time_left_for_this_task=2*60,
      n_jobs=-1,
      metric=autosklearn.metrics.accuracy)

# perform the search
model2.fit(X_train, y_train)

Out[]: AutoSklearnClassifier(metric=accuracy, n_jobs=-1, per_run_time_limit=48,
                      time_left_for_this_task=120)

After about 2 minutes we explore the models.

print(model2.show_models())

[(0.520000, SimpleClassificationPipeline({'balancing:strategy': 'none', 'classifier:__choice__': 'random_forest', 'data_preprocessing:categorical_transformer:categorical_encoding:__choice__': 'one_hot_encoding', 'data_preprocessing:categorical_transformer:category_coalescence:__choice__': 'minority_coalescer', 'data_preprocessing:numerical_transformer:imputation:strategy': 'mean', 'data_preprocessing:numerical_transformer:rescaling:__choice__': 'standardize', 'feature_preprocessor:__choice__': 'no_preprocessing', 'classifier:random_forest:bootstrap': 'True', 'classifier:random_forest:criterion': 'gini', 'classifier:random_forest:max_depth': 'None', 'classifier:random_forest:max_features': 0.5, 'classifier:random_forest:max_leaf_nodes': 'None', 'classifier:random_forest:min_impurity_decrease': 0.0, 'classifier:random_forest:min_samples_leaf': 1, 'classifier:random_forest:min_samples_split': 2, 'classifier:random_forest:min_weight_fraction_leaf': 0.0, 'data_preprocessing:categorical_transformer:category_coalescence:minority_coalescer:minimum_fraction': 0.01},
dataset_properties={
  'task': 1,
  'sparse': False,
  'multilabel': False,
  'multiclass': False,
  'target_type': 'classification',
  'signed': False})),
(0.480000, SimpleClassificationPipeline({'balancing:strategy': 'none', 'classifier:__choice__': 'random_forest', 'data_preprocessing:categorical_transformer:categorical_encoding:__choice__': 'no_encoding', 'data_preprocessing:categorical_transformer:category_coalescence:__choice__': 'minority_coalescer', 'data_preprocessing:numerical_transformer:imputation:strategy': 'most_frequent', 'data_preprocessing:numerical_transformer:rescaling:__choice__': 'standardize', 'feature_preprocessor:__choice__': 'feature_agglomeration', 'classifier:random_forest:bootstrap': 'True', 'classifier:random_forest:criterion': 'entropy', 'classifier:random_forest:max_depth': 'None', 'classifier:random_forest:max_features': 0.48846965177813817, 'classifier:random_forest:max_leaf_nodes': 'None', 'classifier:random_forest:min_impurity_decrease': 0.0, 'classifier:random_forest:min_samples_leaf': 1, 'classifier:random_forest:min_samples_split': 5, 'classifier:random_forest:min_weight_fraction_leaf': 0.0, 'data_preprocessing:categorical_transformer:category_coalescence:minority_coalescer:minimum_fraction': 0.01087424610670389, 'feature_preprocessor:feature_agglomeration:affinity': 'cosine', 'feature_preprocessor:feature_agglomeration:linkage': 'complete', 'feature_preprocessor:feature_agglomeration:n_clusters': 17, 'feature_preprocessor:feature_agglomeration:pooling_func': 'median'},
dataset_properties={
  'task': 1,
  'sparse': False,
  'multilabel': False,
  'multiclass': False,
  'target_type': 'classification',
  'signed': False})),
]

In this particular case it has evaluated two models and we can display some basic statistics about this process.

# summarize
print(model2.sprint_statistics())

auto-sklearn results:
  Dataset name: ecd21bb4-912e-11eb-8af6-acde48001122
  Metric: accuracy
  Best validation score: 0.895218
  Number of target algorithm runs: 12
  Number of successful target algorithm runs: 2
  Number of crashed target algorithm runs: 0
  Number of target algorithms that exceeded the time limit: 10
  Number of target algorithms that exceeded the memory limit: 0

It only had time to create and evaluate 2 models, returning the best model. This can use this model to evaluate results from the holdout test data set.

# evaluate best model
y_predictions = model2.predict(X_test)
acc = accuracy_score(y_test, y_predictions)
print("Accuracy: %.3f" % acc)

Accuracy: 0.900

Now change the run time to see how many extra models will be evaluated in the time. The following increases the run time from 2 to 3 minutes. The evaluation metric has been changed to the f1 score.

# define search
model3 = autosklearn.classification.AutoSklearnClassifier(time_left_for_this_task=3*60,
      n_jobs=4,
      metric=autosklearn.metrics.f1) #accuracy)  #roc_auc   f1)
# perform the search
model3.fit(X_train, y_train)

AutoSklearnClassifier(metric=f1, n_jobs=4, per_run_time_limit=72,
                      time_left_for_this_task=180)

The statistics tells us it evaluated 7 models, out of a target of 15.

# summarize
print(model3.sprint_statistics())

auto-sklearn results:
  Dataset name: 752a4fc6-9135-11eb-8af6-acde48001122
  Metric: f1
  Best validation score: 0.473426
  Number of target algorithm runs: 15
  Number of successful target algorithm runs: 7
  Number of crashed target algorithm runs: 0
  Number of target algorithms that exceeded the time limit: 8
  Number of target algorithms that exceeded the memory limit: 0

The output from the ‘show_models’ function is too long to show here, but you should run it to see the details.

There is a package/library called PipelineProfiler, which is a VERY useful tool for inspecting the various models created and evaluated in the above process. It allows us to see, for each model run, what steps and algorithms were part of it, and by clicking on one we get a flow chart of the pipleline. An example is shown below.

import PipelineProfiler

profiler_data= PipelineProfiler.import_autosklearn(model3)
PipelineProfiler.plot_pipeline_matrix(profiler_data)

OML4Py – AutoML – An Example

Posted on Updated on

OML4Py (Oracle Machine Learning for Python) is Oracle’s offering where you can use Python commands to process and analyse data in an Oracle Database without having to write any SQL. OML4Py, via it’s transparency layer, translates Python code into SQL, executes it in the Database and then presents the results back to you in your Python environment. The examples shown in this post used the OML Notebooks available with Autonomous Databases on Oracle Cloud.

[Warning: the functionality available with initial release of OML4Py is very limited and may not suit most Python developers. Hopefully this will be addressed in later releases]

One of the features of OML4Py is Automated Machine Leaning (AutoML). At some point in the near future Oracle will have a GUI interface for AutoML, which will save you from having to write any code, such as the example in this post. See my previous blog post about AutoML. It is a general discussion on AutoML and some things you need to be careful with. Also, be careful of the marketing around AutoML from all vendors. The reality doesn’t necessarily live up to marketing

OML4Py has a couple of approaches you can follow to Automatically generate a Machine Learning Model (see previous blog post). The first of these can be considered the Black Box approach for AutoML, and the example below illustrates an example of this. The more detailed version of AutoML will be covered in a later post.

[Info: I’m using Oracle Free Tier Database. At time of writing this post OML4Py is only available with Oracle Autonomous 19c]

But before look at these, the first step we need to do is setup the data set to use for AutoML. I’ll be using the popular Portuguese Bank data set. Each code snippets shown below are for a one cell in my OML Notebooks. The data set exists as a table in my schema called BANK_ADDITIONAL_FULL. The sync command creates a proxy object in the notebook session pointing to the table in the DB. No data is copied into the notebook.

%python
import oml
from oml import automl
import pandas as pd
%python
oml_bank = oml.sync(table = 'BANK_ADDITIONAL_FULL')
type(oml_bank)

Let’s explore the data. Remember the data lives in a table in the DB and only the results are displayed

%python
oml_bank.head()

%python
oml_bank.describe()

Now remove one attribute from data set and at the sample time setup the dataframes for input to the ML. This is highly correlated to the the target variable.

%python
oml_bank_X, oml_bank_y = oml_bank.drop('TARGET_Y'), oml_bank['TARGET_Y']

Finally, we can now look at the first of the AutoML options, the black box option. This uses the AutoML ModelSelection function. Using this you can define the type of machine learning to perform (‘classification) and set some additional parameters. The parallel parameter will probably not have too much of an effect when using the Oracle Free Tier, but will certainly improved performance when using additional compute resources.

The example below is very simple and the setup of it is very simple. The ModelSelection function sets up the parameters for the AutoML to function. The ‘select’ function runs the AutoML based on those parameters along with some additional ones. These parameters and the additional ones available are explained below, after this first example.

%python
ms_bank = automl.ModelSelection(mining_function='classification', parallel=4)

ModelSelection can have the following parameters. The possible values for each are listed with the value in bold being the default value:

  • mining_function : the type of ML to preform, only two option available for this,  classification or regression
  • score_metric: what metric to use for evaluating the models. Defaults for binary and multi classification balanced_accuracy is used and default for regression is neg_mean_squared_error. Other options for regression include r2, neg_mean_absolute_error and neg_median_absolute_error.  For classification other options include, accuracy, f1, precision, recall, roc_auc, f1_micro, f1_macro, f1_weighted, recall_micro, recall_macro, recall_weighted, precision_micro, precision_macro, precision_weighted
  • parallel: degree of parallelism to use,  None or a number.

Having defined ModelSelection settings, we can move onto using it to preform (black box) AutoML, using the ‘select’ function. Oracle doesn’t tell us what it does inside this black box except that it uses ML and meta-learning techniques to work out which algorithms to use, what subsets of the original data set to use to give use a optimal outcome. It’s there secret recipe!

The ‘select’ function elevates all the available algorithms, creating models for each or a subset of them based on the meta-learning, and returns the “best” one. The function returns just one model, which is the “best”. The value set for ‘k’ tells the function how many of the “best” or top models created, how many of these to tune before returning the “best” one.

Now, let’s run an example of the ‘select’ function and what parameters is can have

  • X: input data set consisting of the columns to use for Training.
  • y: the column containing the Target variable.
  • case_id: columns name of case_id, default is None. If supplied can be used for data sampling
  • k: the number of (best) models to tune. Default is 3, but can be set to any number between one and eight, as setting it higher than that has no effect as there aren’t any more than that number of algorithms in the database!
  • solver: allowed values are fast (default) and exhaustive. fast uses internal ML and meta-learning thereby reducing the search space.  exhaustive will be slower as it will evaluate all algorithms and options for creating a model.
  • cv: cross validation. Default is auto, but can be set to a number or set to None uses inputs defined in X_valid and y_valid defined below. auto will determine the number based on size of input data set, and when a number is provided will perform that number cross validation.
  • adaptive_sampling: use adaptive sampling to reduce data set size to speed up runtime of ‘select’ function. Default is True, otherwise use False.
  • X_valid: validation data set, default is None.
  • y_valid: validation target column, default is None.
  • time_budget: defines a time constraint on how how long, in seconds, to spend working out the solution. Default is None, or number for number of seconds. Useful for large data sets or for when you need a quicker results, and can be increased based on experimentation.

Here is a basic example of using the ‘select’ function, using the data frames created above as input, ‘k’ is set to five telling the function to tune the top five models created based on doing five fold cross-validation ‘cv’.

best_model = ms_bank.select(oml_bank_X, oml_bank_y, k=5, cv=5) 
best_model

This returns the following model information. We are told the algorithm used (RandomForest), the tuned algorithm settings, and what attributes from the input data frame are used in the tuned model.

(
Algorithm Name: Random Forest

Mining Function: CLASSIFICATION

Target: TARGET_Y

Settings: 
setting name setting value
0 ALGO_NAME ALGO_RANDOM_FOREST
1 CLAS_MAX_SUP_BINS 32
2 CLAS_WEIGHTS_BALANCED OFF
3 ODMS_DETAILS ODMS_DISABLE
4 ODMS_MISSING_VALUE_TREATMENT ODMS_MISSING_VALUE_AUTO
5 ODMS_RANDOM_SEED 0
6 ODMS_SAMPLING ODMS_SAMPLING_DISABLE
7 PREP_AUTO ON
8 RFOR_MTRY 10
9 RFOR_NUM_TREES 20
10 RFOR_SAMPLING_RATIO 0.5
11 TREE_IMPURITY_METRIC TREE_IMPURITY_ENTROPY
12 TREE_TERM_MAX_DEPTH 16
13 TREE_TERM_MINPCT_NODE 0.05
14 TREE_TERM_MINPCT_SPLIT 0.1
15 TREE_TERM_MINREC_NODE 10
16 TREE_TERM_MINREC_SPLIT 20

Attributes: 
AGE
CAMPAIGN
CONS_CONF_IDX
CONS_PRICE_IDX
CONTACT
DEFAULT_VALUE
DURATION
EDUCATION
EMP_VAR_RATE
EURIBOR3M
JOB
MARITAL
MONTH
NR_EMPLOYED
PDAYS
POUTCOME
PREVIOUS

Partition: NO

, 'rf')

[I’ve found the Oracle Documentation for (initial release of) OML4Py lacking with information. Hopefully the documentation will be updated]

I’ve mentioned before you need to exercise some caution with using AutoML due to various potential legal and moral issues. Can they be used as a quick way get an idea if ML will produce useful insights for your data. But the results from it should never be used for making business decisions and never deployed in production. Use it as a starting point, from which to build out an ML solutions with humans making the decisions on what to use and why to use them.

For a more detailed, step-by-step approach to AutoML check out this next post for more.

[Warning: Based on the functionality currently available in this early release of OML4Py, you will be limited in what you can do, not just with AutoML but with other features of OML4Py. Maybe check back at a later time when it has matured and has way more functionality, allowing you to do something useful with it!]

Exploring Database trends using Python pytrends (Google Trends)

Posted on

A little word of warning before you read the rest of this post. The examples shown below are just examples of what is possible. It isn’t very scientific or rigorous, so don’t come complaining if what is shown doesn’t match your knowledge and other insights. This is just a little fun to see what is possible. Yes a more rigorous scientific study is needed, and some attempts at this can be seen at DB-Engines.com. Less scientific are examples shown at TOPDB Top Database index and that isn’t meant to be very scientific.

After all of that, here we go 🙂

pytrends is a library providing an API to Google Trends using Python. The following examples show some ways you can use this library and the focus area I’ll be using is Databases. Many of you are already familiar with using Google Trends, and if this isn’t something you have looked at before then I’d encourage you to go have a look at their website and to give it a try. You don’t need to run Python to use it. For example, here is a quick example taken from the Google Trends website. Here are a couple of screen shots from Google Trends, comparing Relational Database to NoSQL Database. The information presented is based on what searches have been performed over the past 12 months. Some of the information is kind of interesting when you look at the related queries and also the distribution of countries.

To install pytrends use the pip command

pip3 install pytrends

As usual it will change the various pendent libraries and will update where necessary. In my particular case, the only library it updated was the version of pandas.

You do need to be careful of how many searches you perform as you may be limited due to Google rate limits. You can get around this by using a proxy and there is an example on the pytrends PyPi website on how to get around this.

The following code illustrates how to import and setup an initial request. The pandas library is also loaded as the data returned by pytrends API into a pandas dataframe. This will make it ease to format and explore the data.

import pandas as pd 
from pytrends.request import TrendReq

pytrends = TrendReq()

The pytrends API has about nine methods. For my example I’ll be using the following:

  • Interest Over Time: returns historical, indexed data for when the keyword was searched most as shown on Google Trends’ Interest Over Time section.
  • Interest by Region: returns data for where the keyword is most searched as shown on Google Trends’ Interest by Region section.
  • Related Queries: returns data for the related keywords to a provided keyword shown on Google Trends’ Related Queries section.
  • Suggestions: returns a list of additional suggested keywords that can be used to refine a trend search.

Let’s now explore these APIs using the Databases as the main topic of investigation and examining some of the different products. I’ve used the db-engines.com website to select the top 5 databases (as per date of this blog post). These were:

  • Oracle
  • MySQL
  • SQL Server
  • PostgreSQL
  • MongoDB

I will use this list to look for number of searches and other related information. First thing is to import the necessary libraries and create the connection to Google Trends.

import pandas as pd 
from pytrends.request import TrendReq

pytrends = TrendReq()

Next setup the payload and keep the timeframe for searches to the past 12 months only.

search_list = ["Oracle", "MySQL", "SQL Server", "PostgreSQL", "MongoDB"] #max of 5 values allowed
pytrends.build_payload(search_list, timeframe='today 12-m')

We can now look at the the interest over time method to see the number of searches, based on a ranking where 100 is the most popular.

df_ot = pd.DataFrame(pytrends.interest_over_time()).drop(columns='isPartial')
df_ot

and to see a breakdown of these number on an hourly bases you can use the get_historical_interest method.

pytrends.get_historical_interest(search_list)

Let’s move on to exploring the level of interest/searches by country. The following retrieves this information, ordered by Oracle (in decending order) and then select the top 20 countries. Here we can see the relative number of searches per country. Note these doe not necessarily related to the countries with the largest number of searches

df_ibr = pytrends.interest_by_region(resolution='COUNTRY') # CITY, COUNTRY or REGION
df_ibr.sort_values('Oracle', ascending=False).head(20)

Visualizing data is always a good thing to do as we can see a patterns and differences in the data in a clearer way. The following takes the above query and creates a stacked bar chart.

import matplotlib
from matplotlib import pyplot as plt

df2 = df_ibr.sort_values('Oracle', ascending=False).head(20)

df2.reset_index().plot(x='geoName', y=['Oracle', 'MySQL', 'SQL Server', 'PostgreSQL', 'MongoDB'], kind ='bar', stacked=True, title="Searches by Country")

plt.rcParams["figure.figsize"] = [20, 8]
plt.xlabel("Country")
plt.ylabel("Ranking")

We can delve into the data more, by focusing on one particular country and examine the google searches by city or region. The following looks at the data from USA and gives the rankings for the various states.

pytrends.build_payload(search_list, geo='US')
df_ibr = pytrends.interest_by_region(resolution='COUNTRY', inc_low_vol=True)
df_ibr.sort_values('Oracle', ascending=False).head(20)

df2.reset_index().plot(x='geoName', y=['Oracle', 'MySQL', 'SQL Server', 'PostgreSQL', 'MongoDB'], kind ='bar', stacked=True, title="test")
plt.rcParams["figure.figsize"] = [20, 8]

plt.title("Searches for USA")
plt.xlabel("State")
plt.ylabel("Ranking")

 

We can find the top related queries and and top queries including the names of each database.

search_list = ["Oracle", "MySQL", "SQL Server", "PostgreSQL", "MongoDB"] #max of 5 values allowed
pytrends.build_payload(search_list, timeframe='today 12-m')

rq = pytrends.related_queries()
rq.values()

#display rising terms
rq.get('Oracle').get('rising')

We can see the top related rising queries for Oracle are about tik tok. No real surprise there!

and the top queries for Oracle included:

rq.get('Oracle').get('top')

This was an interesting exercise to do. I didn’t show all the results, but when you explore the other databases in the list and see the results from those, and then compare them across the five databases you get to see some interesting patterns.

 

k-Fold and Repeated k-Fold Cross Validation in Python

Posted on Updated on

When it comes to evaluation the performance of a machine learning model there are a number of different approaches. Plus there are as many different view points on what is the best or better evaluation metric to use.

One of the common approaches is to use k-Fold cross validation. This divides the data in to ‘k‘ non-overlapping parts (or Folds). One of these part/Folds is used for hold out testing and the remaining part/Folds (k-1) are used to train and create a model. This model is then used to applied or fitted to the hold-out ‘k‘ part/Fold. This process is repeated across all the ‘k‘ parts/Folds until all the data has been used. The results from applying or fitting the model are aggregated and the mean performance is report.

Traditionally, ‘k‘ is set to 10 and will be the default value in most/all languages, libraries, packages and application. This number can be changed to anything you want. Most reports indicated a value of between 5 and 10, as these seem to indicate results that don’t suffer from bias or variance.

Let’s take a look at an example of using k-Fold Cross Validation using Scikit-Learning library. First step is to prepare the data.

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

bank_file = "/.../4-Datasets/bank-additional-full.csv"

# import dataset
df = pd.read_csv(bank_file, sep=';',)

# get basic details of df (num records, num features)
df.shape

print('Percentage per target class ')
df['y'].value_counts()/len(df) #calculate percentages

#Data Clean up
df = df.drop('duration', axis=1) #this is highly correlated to target variable
df_new = pd.get_dummies(df) #simple and easy approach for categorical variables
df_new.describe()
df['y'] = df['y'].map({'no':0, 'yes':1}) # binary encoding of class label

#split data set into input variables and target variables
## create separate dataframes for Input features (X) and for Target feature (Y)
X_train = df_new.drop('y', axis=1)
Y_train = df_new['y']

Now we can perform k-fold cross valuation.

#load scikit-learn k-fold cross-validation
from numpy import mean
from numpy import std
from sklearn.datasets import make_classification
from sklearn.model_selection import KFold
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression

#setup for k-Fold Cross Validation
cv = KFold(n_splits=10, shuffle=True, random_state=1)
#n_splits = number of k-folds
#shuffle = shuffles data set prior to split
#radnom_state = seed for (pseydo)random number generator
#define model
model = LogisticRegression()
#create model, perform cross validation and evaluate model
scores = cross_val_score(model, X_train, Y_train, scoring='accuracy', cv=cv, n_jobs=-1)
#performance result
print('Accuracy: %.3f (%.3f)' % (mean(scores), std(scores)))

 

We can see from the above example the model is evaluated across 10 folds, giving the accuracy score for each of these. The mean of these 10 accuracy scores is calculated along with the standard deviation, which in this example is very small. You may have slightly different results and this will vary from data set to data set.

The results from k-fold can be nosy, as in each time the code is run a slightly different result may be achieved. This is due to having differing splits of the data set into the k-folds. The model accuracy can vary between each execution and it can be difficult to determine which iteration of the model should be used.

One way to address this possible noise is to estimate the model accurary/performance based on running k-fold a number of times and calculating the performance across all the repeats. This approach is called Repeated k-Fold Cross-Validation. Yes there is a computation cost for performing this approach, and it therefore suited to datasets of smaller scale. In most scenarios having data sets up to 1M records/cases is possible, and depending on the hardware and memory, it can scale to many times that and still be relatively quick to run.

[a small data set for one person could be another persons Big Data set!]

How many repeats should be performed? It kind of depends on how noisy the data is, but in a similar way of having ten as a default value for k, the number of repeats default is ten. Although the typical default is ten, but can be adjusted to say 5, but some testing/experimentation is needed to determine a suitable value.

Building upon the k-fold example code given previously, the following shows can example of using the Repeated k-Fold Cross Validation.

#Repeated k-Fold Cross Validation
#load the necessary libraries
from numpy import mean
from numpy import std
from sklearn.datasets import make_classification
from sklearn.model_selection import RepeatedKFold
from sklearn.model_selection import cross_val_score
from sklearn.linear_model import LogisticRegression

#using the same data set created for k-Fold => X_train, Y_train

#Setup and configure settings for Repeated k-Fold CV (k-folds=10, repeats=10)
rcv = RepeatedKFold(n_splits=10, n_repeats=10, random_state=1)

#define model
model = LogisticRegression()

#create model, perform Repeated CV and evaluate model
scores = cross_val_score(model, X_train, Y_train, scoring='accuracy', cv=rcv, n_jobs=-1)
# report performance
print('Accuracy: %.3f (%.3f)' % (mean(scores), std(scores)))

 

Loading and Reading Binary files in Oracle Database using Python

Posted on Updated on

Most Python example show how to load data into a database table. In this blog post I’ll show you how to load a binary file, for example a picture, into a table in an Oracle Autonomous Database (ATP or ADW) and how to read that same image back using Python.

Before we can do this, we need to setup a few things. These include,

  1. create a table in the database to store the pictures

Let’s use the following table

CREATE TABLE demo_blob (
   id     NUMBER PRIMARY KEY,
   image_txt   VARCHAR2(100),
   image   BLOB);
  1. create a connection to the database

Now let’s get onto the fun bit of loading a image file into this table. The image I’m going to use is the cover of my Data Science book published by MIT Press.

I have this file saved in ‘…/MyBooks/DataScience/BookCover.jpg’.

#Read the binary file
with open (".../MyBooks/DataScience/BookCover.jpg", 'rb') as file:
    blob_file = file.read()

#Display some details of file
print('Length =', len(blob_file))
print('Printing first part of file')
print(blob_file[:50])

Now define the insert statement and setup a cursor to process the insert statement;

#define prepared statement
inst_blob = 'insert into demo_blob (id, image_txt, image) values (:1, :2, :3)'

#connection created using cx_Oracle - see links earlier in post
cur = con.cursor()

Now insert the data and the binary file.

#setup values for attributes
idNum = 1
imageText = 'Demo inserting Blob file'
#insert data into table
cur.execute(inst_blob, (idNum, imageText, blob_file))

#close and finish
cur.close() #close the cursor
con.close() #close the database connection

The image is now saved in the database table. You can use Python to retrieve it or use other tools to view the image.
For example using SQL Developer, query the table and in the results window double click on the blob value. A window pops open and you can view on the image from there by clicking on the check box.

Now that we have the image loads into an Oracle Database the next step is the Python code to read and display the image.

#define prepared statement
qry_blog = 'select id, image_txt, image from demo_blob where id = :1'

#connection created using cx_Oracle - see links earlier in post
cur = con.cursor()
#setup values for attributes 
idNum = 1 
#execute the query
#query the data and blob data
connection.outputtypehandler = OutputTypeHandler
cur.execute(qry_blob, (idNum)) 
id, desc, blob_data = cur.fetchone()

#write the blob data to file
newFileName = '.../MyBooks/DataScience/DummyImage.jpg'
with open(newFileName, 'wb') as file:
   file.write(blob_data)
#close and finish 
cur.close() #close the cursor 
con.close() #close the database connection

Python-Connecting to multiple Oracle Autonomous DBs in one program

Posted on Updated on

More and more people are using the FREE Oracle Autonomous Database for building new new applications, or are migrating to it.

I’ve previously written about connecting to an Oracle Database using Python. Check out that post for details of how to setup Oracle Client and the Oracle Python library cx_Oracle.

In thatblog post I gave examples of connecting to an Oracle Database using the HostName (or IP address), the Service Name or the SID.

But with the Autonomous Oracle Database things are a little bit different. With the Autonomous Oracle Database (ADW or ATP) you will need to use an Oracle Wallet file. This file contains some of the connection details, but you don’t have access to ServiceName/SID, HostName, etc.  Instead you have the name of the Autonomous Database. The Wallet is used to create a secure connection to the Autonomous Database.

You can download the Wallet file from the Database console on Oracle Cloud.

Screenshot 2020-01-10 12.24.10

Most people end up working with multiple database. Sometimes these can be combined into one TNSNAMES file. This can make things simple and easy. To use the download TNSNAME file you will need to set the TNS_ADMIN environment variable. This will allow Python and cx_Oracle library to automatically pick up this file and you can connect to the ATP/ADW Database.

But most people don’t work with just a single database or use a single TNSNAMES file. In most cases you need to switch between different database connections and hence need to use multiple TNSNAMES files.

The question is how can you switch between ATP/ADW Database using different TNSNAMES files while inside one Python program?

Use the os.environ setting in Python. This allows you to reassign the TNS_ADMIN environment variable to point to a new directory containing the TNSNAMES file. This is a temporary assignment and over rides the TNS_ADMIN environment variable.

For example,

import cx_Oracle
import os

os.environ['TNS_ADMIN'] = "/Users/brendan.tierney/Dropbox/wallet_ATP"

p_username = ''p_password = ''p_service = 'atp_high'
con = cx_Oracle.connect(p_username, p_password, p_service)

print(con)
print(con.version)
con.close()

I can now easily switch to another ATP/ADW Database, in the same Python program, by changing the value of os.environ and opening a new connection.

import cx_Oracle
import os

os.environ['TNS_ADMIN'] = "/Users/brendan.tierney/Dropbox/wallet_ATP"
p_username = ''
p_password = ''
p_service = 'atp_high'
con1 = cx_Oracle.connect(p_username, p_password, p_service)
...
con1.close()

...
os.environ['TNS_ADMIN'] = "/Users/brendan.tierney/Dropbox/wallet_ADW2"
p_username = ''
p_password = ''
p_service = 'ADW2_high'
con2 = cx_Oracle.connect(p_username, p_password, p_service)
...
con2.close()

As mentioned previously the setting and resetting of TNS_ADMIN using os.environ, is only temporary, and when your Python program exists or completes the original value for this environment variable will remain.

#GE2020 Comparing Party Manifestos to 2016

Posted on

A few days ago I wrote a blog post about using Python to analyze the 2016 general (government) elections manifestos of the four main political parties in Ireland.

Today the two (traditional) largest parties released their #GE2020 manifestos. You can get them by following these links

The following images show the WordClouds generated for the #GE2020 Manifestos. I used the same Python code used in my previous post. If you want to try this out yourself, all the Python code is there.

First let us look at the WordClouds from Fine Gael.

FG2020
2020 Manifesto

FG_2016
2016 Manifesto

Now for the Fianna Fail WordClouds.

FF2020
2020 Manifesto

FF_2016
2016 Manifesto

When you look closely at the differences between the manifestos you will notice there are some common themes across the manifestos from 2016 to those in the 2020 manifestos. It is also interesting to see some new words appearing/disappearing for the 2020 manifestos. Some of these are a little surprising and interesting.

#GE2020 Analysing Party Manifestos using Python

Posted on

The general election is underway here in Ireland with polling day set for Saturday 8th February. All the politicians are out campaigning and every day the various parties are looking for publicity on whatever the popular topic is for that day. Each day is it a different topic.

Most of the political parties have not released their manifestos for the #GE2020 election (as of date of this post). I want to use some simple Python code to perform some analyse of their manifestos. As their new manifestos weren’t available (yet) I went looking for their manifestos from the previous general election. Michael Pidgeon has a website with party manifestos dating back to the early 1970s, and also has some from earlier elections. Check out his website.

I decided to look at manifestos from the 4 main political parties from the 2016 general election. Yes there are other manifestos available, and you can use the Python code, given below to analyse those, with only some minor edits required.

The end result of this simple analyse is a WordCloud showing the most commonly used words in their manifestos. This is graphical way to see what some of the main themes and emphasis are for each party, and also allows us to see some commonality between the parties.

Let’s begin with the Python code.

1 – Initial Setup

There are a number of Python Libraries available for processing PDF files. Not all of them worked on all of the Part Manifestos PDFs! It kind of depends on how these files were generated. In my case I used the pdfminer library, as it worked with all four manifestos. The common library PyPDF2 didn’t work with the Fine Gael manifesto document.

import io
import pdfminer
from pprint import pprint
from pdfminer.converter import TextConverter
from pdfminer.pdfinterp import PDFPageInterpreter
from pdfminer.pdfinterp import PDFResourceManager
from pdfminer.pdfpage import PDFPage

#directory were manifestos are located
wkDir = '.../General_Election_Ire/'

#define the names of the Manifesto PDF files & setup party flag
pdfFile = wkDir+'FGManifesto16_2.pdf'
party = 'FG'
#pdfFile = wkDir+'Fianna_Fail_GE_2016.pdf'
#party = 'FF'
#pdfFile = wkDir+'Labour_GE_2016.pdf'
#party = 'LB'
#pdfFile = wkDir+'Sinn_Fein_GE_2016.pdf'
#party = 'SF'

All of the following code will run for a given manifesto. Just comment in or out the manifesto you are interested in. The WordClouds for each are given below.

2 – Load the PDF File into Python

The following code loops through each page in the PDF file and extracts the text from that page.

I added some addition code to ignore pages containing the Irish Language. The Sinn Fein Manifesto contained a number of pages which were the Irish equivalent of the preceding pages in English. I didn’t want to have a mixture of languages in the final output.

SF_IrishPages = [14,15,16,17,18,19,20,21,22,23,24]
text = ""

pageCounter = 0
resource_manager = PDFResourceManager()
fake_file_handle = io.StringIO()
converter = TextConverter(resource_manager, fake_file_handle)
page_interpreter = PDFPageInterpreter(resource_manager, converter)

for page in PDFPage.get_pages(open(pdfFile,'rb'), caching=True, check_extractable=True):
    if (party == 'SF') and (pageCounter in SF_IrishPages):
        print(party+' - Not extracting page - Irish page', pageCounter)
    else:
        print(party+' - Extracting Page text', pageCounter)
        page_interpreter.process_page(page)

        text = fake_file_handle.getvalue()

    pageCounter += 1

print('Finished processing PDF document')
converter.close()
fake_file_handle.close()
FG - Extracting Page text 0
FG - Extracting Page text 1
FG - Extracting Page text 2
FG - Extracting Page text 3
FG - Extracting Page text 4
FG - Extracting Page text 5
...

3 – Tokenize the Words

The next step is to Tokenize the text. This breaks the text into individual words.

from nltk.tokenize import word_tokenize
from nltk.corpus import stopwords
tokens = []

tokens = word_tokenize(text)

print('Number of Pages =', pageCounter)
print('Number of Tokens =',len(tokens))
Number of Pages = 140
Number of Tokens = 66975

4 – Filter words, Remove Numbers & Punctuation

There will be a lot of things in the text that we don’t want included in the analyse. We want the text to only contain words. The following extracts the words and ignores numbers, punctuation, etc.

#converts to lower case, and removes punctuation and numbers
wordsFiltered = [tokens.lower() for tokens in tokens if tokens.isalpha()]
print(len(wordsFiltered))
print(wordsFiltered)
58198
['fine', 'gael', 'general', 'election', 'manifesto', 's', 'keep', 'the', 'recovery', 'going', 'gaelgeneral', 'election', 'manifesto', 'foreward', 'from', 'an', 'taoiseach', 'the', 'long', 'term', 'economic', 'three', 'steps', 'to', 'keep', 'the', 'recovery', 'going', 'agriculture', 'and', 'food', 'generational',
...

As you can see the number of tokens has reduced from 66,975 to 58,198.

5 – Setup Stop Words

Stop words are general words in a language that doesn’t contain any meanings and these can be removed from the data set. Python NLTK comes with a set of stop words defined for most languages.

#We initialize the stopwords variable which is a list of words like 
#"The", "I", "and", etc. that don't hold much value as keywords
stop_words = stopwords.words('english')
print(stop_words)
['i', 'me', 'my', 'myself', 'we', 'our', 'ours', 'ourselves', 'you', "you're", "you've", "you'll", "you'd", 'your', 'yours', 'yourself',
....

Additional stop words can be added to this list. I added the words listed below. Some of these you might expect to be in the stop word list, others are to remove certain words that appeared in the various manifestos that don’t have a lot of meaning. I also added the name of the parties  and some Irish words to the stop words list.

#some extra stop words are needed after examining the data and word cloud
#these are added
extra_stop_words = ['ireland','irish','ł','need', 'also', 'set', 'within', 'use', 'order', 'would', 'year', 'per', 'time', 'place', 'must', 'years', 'much', 'take','make','making','manifesto','ð','u','part','needs','next','keep','election', 'fine','gael', 'gaelgeneral', 'fianna', 'fáil','fail','labour', 'sinn', 'fein','féin','atá','go','le','ar','agus','na','ár','ag','haghaidh','téarnamh','bplean','page','two','number','cothromfor']
stop_words.extend(extra_stop_words)
print(stop_words)

Now remove these stop words from the list of tokens.

# remove stop words from tokenised data set
filtered_words = [word for word in wordsFiltered if word not in stop_words]
print(len(filtered_words))
print(filtered_words)
31038
['general', 'recovery', 'going', 'foreward', 'taoiseach', 'long', 'term', 'economic', 'three', 'steps', 'recovery', 'going', 'agriculture', 'food',

The number of tokens is reduced to 31,038

6 – Word Frequency Counts

Now calculate how frequently these words occur in the list of tokens.

#get the frequency of each word
from collections import Counter

# count frequencies
cnt = Counter()
for word in filtered_words:
cnt[word] += 1

print(cnt)
Counter({'new': 340, 'support': 249, 'work': 190, 'public': 186, 'government': 177, 'ensure': 177, 'plan': 176, 'continue': 168, 'local': 150, 
...

7 – WordCloud

We can use the word frequency counts to add emphasis to the WordCloud. The more frequently it occurs the larger it will appear in the WordCloud.

#create a word cloud using frequencies for emphasis 
from wordcloud import WordCloud
import matplotlib.pyplot as plt

wc = WordCloud(max_words=100, margin=9, background_color='white',
scale=3, relative_scaling = 0.5, width=500, height=400,
random_state=1).generate_from_frequencies(cnt)

plt.figure(figsize=(20,10))
plt.imshow(wc)
#plt.axis("off")
plt.show()

#Save the image in the img folder:
wc.to_file(wkDir+party+"_2016.png")

The last line of code saves the WordCloud image as a file in the directory where the manifestos are located.

8 – WordClouds for Each Party

Screenshot 2020-01-21 11.10.25

Remember these WordClouds are for the manifestos from the 2016 general election.

When the parties have released their manifestos for the 2020 general election, I’ll run them through this code and produce the WordClouds for 2020. It will be interesting to see the differences between the 2016 and 2020 manifesto WordClouds.

Data Profiling in Python

Posted on Updated on

With every data analytics and data science project, one of the first tasks to that everyone needs to do is to profile the data sets. Data profiling allows you to get an initial picture of the data set, see data distributions and relationships. Additionally it allows us to see what kind of data cleaning and data transformations are necessary.

Most data analytics tools and languages have some functionality available to help you. Particular the various data science/machine learning products have this functionality built-in them and can do a lot of the data profiling automatically for you. But if you don’t use these tools/products, then you are probably using R and/or Python to profile your data.

With Python you will be working with the data set loaded into a Pandas data frame. From there you will be using various statistical functions and graphing functions (and libraries) to create a data profile. From there you will probably create a data profile report.

But one of the challenges with doing this in Python is having different coding for handling numeric and character based attributes/features. The describe function in Python (similar to the summary function in R) gives some statistical summaries for numeric attributes/features. A different set of functions are needed for character based attributes. The Python Library repository (https://pypi.org/) contains over 200K projects. But which ones are really useful and will help with your data science projects. Especially with new projects and libraries being released on a continual basis? This is a major challenge to know what is new and useful.

For example the followings shows loading the titanic data set into a Pandas data frame, creating a subset and using the describe function in Python.

import pandas as pd

df = pd.read_csv("/Users/brendan.tierney/Dropbox/4-Datasets/titanic/train.csv")

df.head(5)

Screenshot 2019-11-22 16.58.39

df2 = df.iloc[:,[1,2,4,5,6,7,8,10,11]]
df2.head(5)

Screenshot 2019-11-22 16.59.30

df2.describe()

Screenshot 2019-11-22 17.00.17

You will notice the describe function has only looked at the numeric attributes.

One of those 200+k Python libraries is one called pandas_profiling. This will create a data audit report for both numeric and character based attributes. This most be good, Right?  Let’s take a look at what it does.

For each column the following statistics – if relevant for the column type – are presented in an interactive HTML report:

  • Essentials: type, unique values, missing values
  • Quantile statistics like minimum value, Q1, median, Q3, maximum, range, interquartile range
  • Descriptive statistics like mean, mode, standard deviation, sum, median absolute deviation, coefficient of variation, kurtosis, skewness
  • Most frequent values
  • Histogram
  • Correlations highlighting of highly correlated variables, Spearman, Pearson and Kendall matrices
  • Missing values matrix, count, heatmap and dendrogram of missing values

The first step is to install the pandas_profiling library.

pip3 install pandas_profiling

Now run the pandas_profiling report for same data frame created and used, see above.

import pandas_profiling as pp

df2.profile_report()

The following images show screen shots of each part of the report. Click and zoom into these to see more details.

Screenshot 2019-11-22 17.29.00Screenshot 2019-11-22 17.29.46

Screenshot 2019-11-22 17.30.57Screenshot 2019-11-22 17.31.32

Screenshot 2019-11-22 17.31.57Screenshot 2019-11-22 17.32.31

Screenshot 2019-11-22 17.33.02

 

Managing imbalanced Data Sets with SMOTE in Python

Posted on Updated on

When working with data sets for machine learning, lots of these data sets and examples we see have approximately the same number of case records for each of the possible predicted values. In this kind of scenario we are trying to perform some kind of classification, where the machine learning model looks to build a model based on the input data set against a target variable. It is this target variable that contains the value to be predicted. In most cases this target variable (or feature) will contain binary values or equivalent in categorical form such as Yes and No, or A and B, etc or may contain a small number of other possible values (e.g. A, B, C, D).

For the classification algorithm to perform optimally and be able to predict the possible value for a new case record, it will need to see enough case records for each of the possible values. What this means, it would be good to have approximately the same number of records for each value (there are many ways to overcome this and these are outside the score of this post). But most data sets, and those that you will encounter in real life work scenarios, are never balanced, as in having a 50-50 split. What we typically encounter might be a 90-10, 98-2, etc type of split. These data sets are said to be imbalanced.

Screenshot 2019-05-20 15.34.14

The image above gives examples of two approaches for creating a balanced data set. The first is under-sampling. This involves reducing the class that contains the majority of the case records and reducing it to match the number of case records in the minor class. The problems with this include, the resulting data set is too small to be meaningful, the case records removed could contain important records and scenarios that the model will need to know about.

The second example is creating a balanced data set by increasing the number of records in the minority class. There are a few approaches to creating this. The first approach is to create duplicate records, from the minor class, until such time as the number of case records are approximately the same for each class. This is the simplest approach. The second approach is to create synthetic records that are statistically equivalent of the original data set. A commonly technique used for this is called SMOTE, Synthetic Minority Oversampling Technique. SMOTE uses a nearest neighbors algorithm to generate new and synthetic data we can use for training our model. But one of the issues with SMOTE is that it will not create sample records outside the bounds of the original data set. As you can image this would be very difficult to do.

The following examples will illustrate how to perform Under-Sampling and Over-Sampling (duplication and using SMOTE) in Python using functions from Pandas, Imbalanced-Learn and Sci-Kit Learn libraries.

NOTE: The Imbalanced-Learn library (e.g. SMOTE)requires the data to be in numeric format, as it statistical calculations are performed on these. The python function get_dummies was used as a quick and simple to generate the numeric values. Although this is perhaps not the best method to use in a real project. With the other sampling functions can process data sets with a sting and numeric.

Data Set:  Is the Portuaguese Banking data set and is available on the UCI Data Set Repository, and many other sites. Here are some basics with that data set.

import warnings

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
get_ipython().magic('matplotlib inline')

bank_file = ".../bank-additional-full.csv"

# import dataset
df = pd.read_csv(bank_file, sep=';',)

# get basic details of df (num records, num features)
df.shape
df['y'].value_counts() # dataset is imbalanced with majority of class label as "no".
no     36548
yes     4640
Name: y, dtype: int64
#print bar chart
df.y.value_counts().plot(kind='bar', title='Count (target)');

Screenshot 2019-05-27 09.38.36

Example 1a – Down/Under sampling the majority class y=1 (using random sampling)

count_class_0, count_class_1 = df.y.value_counts()

# Divide by class
df_class_0 = df[df['y'] == 0] #majority class
df_class_1 = df[df['y'] == 1] #minority class

# Sample Majority class (y=0, to have same number of records as minority calls (y=1)
df_class_0_under = df_class_0.sample(count_class_1)
# join the dataframes containing y=1 and y=0
df_test_under = pd.concat([df_class_0_under, df_class_1])

print('Random under-sampling:')
print(df_test_under.y.value_counts())
print("Num records = ", df_test_under.shape[0])

df_test_under.y.value_counts().plot(kind='bar', title='Count (target)');
Random under-sampling:
1    4640
0    4640
Name: y, dtype: int64
Num records =  9280

Screenshot 2019-05-27 09.41.06

Example 1b – Down/Under sampling the majority class y=1 using imblearn

from imblearn.under_sampling import RandomUnderSampler

X = df_new.drop('y', axis=1)
Y = df_new['y']

rus = RandomUnderSampler(random_state=42, replacement=True)
X_rus, Y_rus = rus.fit_resample(X, Y)

df_rus = pd.concat([pd.DataFrame(X_rus), pd.DataFrame(Y_rus, columns=['y'])], axis=1)

print('imblearn over-sampling:')
print(df_rus.y.value_counts())
print("Num records = ", df_rus.shape[0])

df_rus.y.value_counts().plot(kind='bar', title='Count (target)');

[same results as Example 1a]

Example 1c – Down/Under sampling the majority class y=1 using Sci-Kit Learn

from sklearn.utils import resample

print("Original Data distribution")
print(df['y'].value_counts())

# Down Sample Majority class
down_sample = resample(df[df['y']==0],
replace = True, # sample with replacement
n_samples = df[df['y']==1].shape[0], # to match minority class
random_state=42) # reproducible results

# Combine majority class with upsampled minority class
train_downsample = pd.concat([df[df['y']==1], down_sample])

# Display new class counts
print('Sci-Kit Learn : resample : Down Sampled data set')
print(train_downsample['y'].value_counts())
print("Num records = ", train_downsample.shape[0])
train_downsample.y.value_counts().plot(kind='bar', title='Count (target)');

[same results as Example 1a]

Example 2 a – Over sampling the minority call y=0 (using random sampling)

df_class_1_over = df_class_1.sample(count_class_0, replace=True)
df_test_over = pd.concat([df_class_0, df_class_1_over], axis=0)

print('Random over-sampling:')
print(df_test_over.y.value_counts())

df_test_over.y.value_counts().plot(kind='bar', title='Count (target)');
Random over-sampling:
1    36548
0    36548
Name: y, dtype: int64

Screenshot 2019-05-27 09.46.08

Example 2b – Over sampling the minority call y=0 using SMOTE

from imblearn.over_sampling import SMOTE

print(df_new.y.value_counts())
X = df_new.drop('y', axis=1)
Y = df_new['y']

sm = SMOTE(random_state=42)
X_res, Y_res = sm.fit_resample(X, Y)

df_smote_over = pd.concat([pd.DataFrame(X_res), pd.DataFrame(Y_res, columns=['y'])], axis=1)

print('SMOTE over-sampling:')
print(df_smote_over.y.value_counts())

df_smote_over.y.value_counts().plot(kind='bar', title='Count (target)');

[same results as Example 2a]

Example 2c – Over sampling the minority call y=0 using Sci-Kit Learn

from sklearn.utils import resample

print("Original Data distribution")
print(df['y'].value_counts())

# Upsample minority class
train_positive_upsample = resample(df[df['y']==1],
replace = True, # sample with replacement
n_samples = train_zero.shape[0], # to match majority class
random_state=42) # reproducible results

# Combine majority class with upsampled minority class
train_upsample = pd.concat([train_negative, train_positive_upsample])

# Display new class counts
print('Sci-Kit Learn : resample : Up Sampled data set')
print(train_upsample['y'].value_counts())
train_upsample.y.value_counts().plot(kind='bar', title='Count (target)');

[same results as Example 2a]

Migrating Python ML Models to other languages

Posted on Updated on

I’ve mentioned in a previous blog post about experiencing some performance issues with using Python ML in production. We needed something quicker and the possible languages we considered were C, C++, Java and Go Lang.

But the data science team used R and Python, with just a few more people using Python than R on the team.

One option was to rewrite everything into the language used in production. As you can imagine no-one wanted to do that and there was no way of ensure a bug free solution and one that gave similar results to the R and Python models. The other option was to look for some code to convert the models from one language to another.

The R users was well versed in using PMML. Predictive Model Markup Language (PMML) has been around a long time and well known and used by certain groups of data scientists who have been around a while. It is also widely supported by many analytics vendors, and provides an inter-change format to allow predictive models to be described and exchanged. For newer people, they hadn’t heard of it. PMML is an XML based interchange specification.

But with PMML there are some limitation. Not with the specification but how it is implemented by the various vendors that support it. PMML supports the exchange of the model pipeline including the data transformations as well as the model specification. Most vendors only support some elements of this and maybe just a couple of models. And there-in lies the problem. How can a ML pipeline be migrated from, as Python, to some other language and/or tool. There are limitations.

If you do want to explore PMML with Python check out the sklearn2pmml package and is also available on PyPl. This package allows you to export the ML pipeline and the model specification. As with most other implementations of PMML there are some parts of the PMML specification not implement, but it is better than post of the other implementation out there.

An alternative is to look at code translations options. With these we want something that will take our ML pipeline and convert it to another programming language like C++, JAVA, Go, etc. There aren’t too many solutions available to do this. One such solution we’ve explored over the past couple of weeks is called m2cgen.

m2cgen (Model 2 Code Generator) is a lightweight library which provides an easy way to transpile trained statistical models into a native code (Python, C, Java, Go). You can supply M2cgen with a range of models (linear, SVM, tree, random forest, or boosting, etc) and the tool will output code in the chosen language that will represent the trained model. The code generated will generated into native code without dependencies. Other packages or libraries are not dependent or required in the translated language. For example here is an example Decision Tree translated into a number of different languages.

 

C

#include <string.h>
void score(double * input, double * output) {
    double var0[3];
    if ((input[2]) <= (2.6)) {
        memcpy(var0, (double[]){1.0, 0.0, 0.0}, 3 * sizeof(double));
    } else {
        if ((input[2]) <= (4.8500004)) {
            if ((input[3]) <= (1.6500001)) {
                memcpy(var0, (double[]){0.0, 1.0, 0.0}, 3 * sizeof(double));
            } else {
                memcpy(var0, (double[]){0.0, 0.3333333333333333, 0.6666666666666666}, 3 * sizeof(double));
            }
        } else {
            if ((input[3]) <= (1.75)) {
                memcpy(var0, (double[]){0.0, 0.42857142857142855, 0.5714285714285714}, 3 * sizeof(double));
            } else {
                memcpy(var0, (double[]){0.0, 0.0, 1.0}, 3 * sizeof(double));
            }
        }
    }
    memcpy(output, var0, 3 * sizeof(double));
}

Java

public class Model {

    public static double[] score(double[] input) {
        double[] var0;
        if ((input[2]) <= (2.6)) {
            var0 = new double[] {1.0, 0.0, 0.0};
        } else {
            if ((input[2]) <= (4.8500004)) {
                if ((input[3]) <= (1.6500001)) {
                    var0 = new double[] {0.0, 1.0, 0.0};
                } else {
                    var0 = new double[] {0.0, 0.3333333333333333, 0.6666666666666666};
                }
            } else {
                if ((input[3]) <= (1.75)) {
                    var0 = new double[] {0.0, 0.42857142857142855, 0.5714285714285714};
                } else {
                    var0 = new double[] {0.0, 0.0, 1.0};
                }
            }
        }
        return var0;
    }
}

Go Lang

func score(input []float64) []float64 {
    var var0 []float64
    if (input[2]) <= (2.6) {
        var0 = []float64{1.0, 0.0, 0.0}
    } else {
        if (input[2]) <= (4.8500004) {
            if (input[3]) <= (1.6500001) {
                var0 = []float64{0.0, 1.0, 0.0}
            } else {
                var0 = []float64{0.0, 0.3333333333333333, 0.6666666666666666}
            }
        } else {
            if (input[3]) <= (1.75) {
                var0 = []float64{0.0, 0.42857142857142855, 0.5714285714285714}
            } else {
                var0 = []float64{0.0, 0.0, 1.0}
            }
        }
    }
    return var0
}