Upgini Intelligent data search & enrichment for Machine Learning and AI

Easily find and add relevant features to your ML & AI pipeline from
hundreds of public, community and premium external data sources,
including open & commercial LLMs

Quick Start in Colab | Register / Sign In | Slack Community | Propose new Data source

Overview

Upgini is an intelligent data search engine with a Python library that helps you find and add relevant features to your ML pipeline from hundreds of public, community, and premium external data sources. Under the hood, Upgini automatically optimizes all connected data sources by generating an optimal set of machine ML features using large language models (LLMs), GraphNNs and recurrent neural networks (RNNs).

Motivation: for most supervised ML models external data & features boost accuracy significantly better than any hyperparameters tuning. But lack of automated and time-efficient enrichment tools for external data blocks massive adoption of external features in ML pipelines. We want radically simplify features search and enrichment to make external data a standard approach. Like a hyperparameter tuning for machine learning nowadays.

Mission: Democratize access to data sources for data science community.

Awesome features

Automatically find only relevant features that give accuracy improvement for ML model. Not just correlated with target variable, what 9 out of 10 cases gives zero accuracy improvement
Automated feature generation from the sources: feature generation with Large Language Models' data augmentation, RNNs, GraphNN; multiple data source ensembling
Automatic search key augmentation from all connected sources. If you do not have all search keys in your search request, such as postal/zip code, Upgini will try to add those keys based on the provided set of search keys. This will broaden the search across all available data sources
Calculate accuracy metrics and uplifts after enrichment existing ML model with external features
Check the stability of accuracy gain from external data on out-of-time intervals and verification datasets. Mitigate risks of unstable external data dependencies in ML pipeline
Easy to use - single request to enrich training dataset with all of the keys at once:

Scikit-learn compatible interface for quick data integration with existing ML pipelines
Support for most common supervised ML tasks on tabular data:

Simple Drag & Drop Search UI:

Connected data sources and coverage

• Public data : public sector, academic institutions, other sources through open data portals. Curated and updated by the Upgini team
  
• Community shared data: royalty / license free datasets or features from Data science community (our users). It's both a public and a scraped data
  
• Premium data providers: commercial data sources verified by the Upgini team in real-world use cases
  

Details on datasets and features

Total: 239 countries and up to 41 years of history

|Data sources|Countries|History, years|# sources for ensemble|Update|Search keys|API Key required |--|--|--|--|--|--|--| |Historical weather & Climate normals | 68 |22|-|Monthly|date, country, postal/ZIP code|No |Location/Places/POI/Area/Proximity information from OpenStreetMap | 221 |2|-|Monthly|date, country, postal/ZIP code|No |International holidays & events, Workweek calendar| 232 |22|-|Monthly|date, country|No |Consumer Confidence index| 44 |22|-|Monthly|date, country|No |World economic indicators|191 |41|-|Monthly|date, country|No |Markets data|-|17|-|Monthly|date, datetime|No |World mobile & fixed broadband network coverage and performance |167|-|3|Monthly|country, postal/ZIP code|No |World demographic data |90|-|2|Annual|country, postal/ZIP code|No |World house prices |44|-|3|Annual|country, postal/ZIP code|No |Public social media profile data |104|-|-|Monthly|date, email/HEM, phone |Yes |Car ownership data and Parking statistics|3|-|-|Annual|country, postal/ZIP code, email/HEM, phone|Yes |Geolocation profile for phone & IPv4 & email|239|-|6|Monthly|date, email/HEM, phone, IPv4|Yes | Email/WWW domain profile|-|-|-|-

Know other useful data sources for machine learning? Give us a hint and we'll add it for free.

Tutorials

Search of relevant external features & Automated feature generation for Salary prediction task (use as a template)

• The goal is to predict salary for data science job postning based on information about employer and job description.
• Following this guide, you'll learn how to search & auto generate new relevant features with Upgini library
• The evaluation metric is Mean Absolute Error (MAE).

Run Feature search & generation notebook inside your browser:

  <!--   -->

Simple sales prediction for retail stores

• The goal is to predict future sales of different goods in stores based on a 5-year history of sales.
  
• Kaggle Competition Store Item Demand Forecasting Challenge is a product sales forecasting. The evaluation metric is SMAPE.

Run Simple sales prediction for retail stores inside your browser:

  <!--   -->

How to boost ML model accuracy for Kaggle TOP1 Leaderboard in 10 minutes

• The goal is accuracy improvement for TOP1 winning Kaggle solution from new relevant external features & data.
  
• Kaggle Competition is a product sales forecasting, evaluation metric is SMAPE.
  

How to do low-code feature engineering for AutoML tools

• Save time on feature search and engineering. Use ready-to-use external features and data sources to maximize overall AutoML accuracy, right out of the box.
  
• Kaggle Competition is a product sales forecasting, evaluation metric is SMAPE.
  
• Low-code AutoML tools: Upgini and PyCaret

How to improve accuracy of Multivariate Time Series forecast from external features & data

• The goal is accuracy improvement of Multivariate Time Series prediction from new relevant external features & data. The main challenge here is a strategy of data & feature enrichment, when a component of Multivariate TS depends not only on its past values but also has some dependency on other components.
  
• Kaggle Competition is a product sales forecasting, evaluation metric is RMSLE.

How to speed up feature engineering hypothesis tests with ready-to-use external features

• Save time on external data wrangling and feature calculation code for hypothesis tests. The key challenge here is a time-dependent representation of information in a training dataset, which is uncommon for credit default prediction tasks. As a result, special data enrichment strategy is used.
  
• Kaggle Competition is a credit default prediction, evaluation metric is normalized Gini coefficient.

Quick start

1. Install from PyPI

python %pip install upgini

2. Use your labeled training dataset for search

You can use your labeled training datasets "as is" to initiate the search. Under the hood, we'll search for relevant data using: - search keys from training dataset to match records from potential data sources with a new features - labels from training dataset to estimate relevancy of feature or dataset for your ML task and calculate feature importance metrics
- your features from training dataset to find external datasets and features which only give accuracy improvement to your existing data and estimate accuracy uplift (optional)

Load training dataset into pandas dataframe and separate features' columns from label column in a Scikit-learn way:
```python import pandas as pd

labeled training dataset - customerchurnprediction_train.csv

traindf = pd.readcsv("customerchurnpredictiontrain.csv") X = traindf.drop(columns="churnflag") y = traindf["churn_flag"] ```

3. Choose one or multiple columns as a search keys

Search keys columns will be used to match records from all potential external data sources / features.
Define one or multiple columns as a search keys with FeaturesEnricher class initialization.
```python from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey

enricher = FeaturesEnricher( searchkeys={ "subscriptionactivationdate": SearchKey.DATE, "country": SearchKey.COUNTRY, "zipcode": SearchKey.POSTALCODE, "hashedemail": SearchKey.HEM, "lastvisitipaddress": SearchKey.IP, "registeredwith_phone": SearchKey.PHONE }) ```

Search key types we support (more to come!)

For the meaning types SearchKey.DATE/SearchKey.DATETIME with dtypes object or string you have to clarify date/datetime format by passing dateformat parameter to FeaturesEnricher. For example: ```python from upgini.featuresenricher import FeaturesEnricher from upgini.metadata import SearchKey

enricher = FeaturesEnricher( searchkeys={ "subscriptionactivationdate": SearchKey.DATE, "country": SearchKey.COUNTRY, "zipcode": SearchKey.POSTALCODE, "hashedemail": SearchKey.HEM, "lastvisitipaddress": SearchKey.IP, "registeredwithphone": SearchKey.PHONE }, dateformat = "%Y-%d-%m" ) ```

To use datetime not in UTC timezone, you can cast datetime column explicitly to your timezone (example for Warsaw): python df["date"] = df.date.astype("datetime64").dt.tz_localize("Europe/Warsaw")

Single country for the whole training dataset can be passed with country_code parameter: ```python from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey

enricher = FeaturesEnricher( searchkeys={ "subscriptionactivationdate": SearchKey.DATE, "zipcode": SearchKey.POSTALCODE, }, countrycode = "US", date_format = "%Y-%d-%m" ) ```

4. Start your first feature search!

The main abstraction you interact is FeaturesEnricher, a Scikit-learn compatible estimator. You can easily add it into your existing ML pipelines. Create instance of the FeaturesEnricher class and call: - fit to search relevant datasets & features
- than transform to enrich your dataset with features from search result

Let's try it out! ```python import pandas as pd from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey

load labeled training dataset to initiate search

traindf = pd.readcsv("customerchurnpredictiontrain.csv") X = traindf.drop(columns="churnflag") y = traindf["churn_flag"]

now we're going to create FeaturesEnricher class

enricher = FeaturesEnricher( searchkeys={ "subscriptionactivationdate": SearchKey.DATE, "country": SearchKey.COUNTRY, "zipcode": SearchKey.POSTAL_CODE })

everything is ready to fit! For 200 records fitting should take around 10 minutes,

we send email notification, just register on profile.upgini.com

enricher.fit(X, y) ```

That's all! We've fit FeaturesEnricher.

5. Evaluate feature importances (SHAP values) from the search result

FeaturesEnricher class has two properties for feature importances, which will be filled after fit - feature_names_ and feature_importances_:
- feature_names_ - feature names from the search result, and if parameter keep_input=True was used, initial columns from search dataset as well
- feature_importances_ - SHAP values for features from the search result, same order as in feature_names_

Method get_features_info() returns pandas dataframe with features and full statistics after fit, including SHAP values and match rates: python enricher.get_features_info() Get more details about FeaturesEnricher at runtime using docstrings via help(FeaturesEnricher) or help(FeaturesEnricher.fit).

6. Enrich Production ML pipeline with relevant external features

FeaturesEnricher is a Scikit-learn compatible estimator, so any pandas dataframe can be enriched with external features from a search result (after fit ).
Use transform method of FeaturesEnricher , and let magic to do the rest ```python

load dataset for enrichment

testx = pd.readcsv("test.csv")

enrich it!

enrichedtestfeatures = enricher.transform(test_x) ``` #### 6.1 Reuse completed search for enrichment without 'fit' run

FeaturesEnricher can be initiated with a search_id parameter from completed search after fit method call.
Just use enricher.get_search_id() or copy search id string from the fit() output.
Search keys and features in X should be the same as for fit() python enricher = FeaturesEnricher( #same set of a search keys as for the fit step search_keys={"date": SearchKey.DATE}, api_key="<YOUR API_KEY>", # if you fit enricher with api_key then you should use it here search_id = "abcdef00-0000-0000-0000-999999999999" ) enriched_prod_dataframe=enricher.transform(input_dataframe)

6.2 Enrichment with an updated external data sources and features

For most of the ML cases, training step requires labeled dataset with a historical observations from the past. But for production step you'll need an updated and actual data sources and features for the present time, to calculate a prediction.
FeaturesEnricher, when initiated with set of search keys which includes SearchKey.DATE, will match records from all potential external data sources exactly on a the specific date/datetime based on SearchKey.DATE. To avoid enrichment with features "form the future" for the fit step.
And then, for transform in a production ML pipeline, you'll get enrichment with relevant features, actual for the present date.

Initiate FeaturesEnricher with SearchKey.DATE search key in a key set to get actual features for production and avoid features from the future for the training: python enricher = FeaturesEnricher( search_keys={ "subscription_activation_date": SearchKey.DATE, "country": SearchKey.COUNTRY, "zip_code": SearchKey.POSTAL_CODE, }, )

How does it work?

Search dataset validation

We validate and clean search initialization dataset under the hood:

• heck you search keys columns format;
• check zero variance for label column;
• check dataset for full row duplicates. If we find any, we remove duplicated rows and make a note on share of row duplicates;
  
• check inconsistent labels - rows with the same features and keys but different labels, we remove them and make a note on share of row duplicates;
  
• remove columns with zero variance - we treat any non search key column in search dataset as a feature, so columns with zero variance will be removed

Supervised ML tasks detection

We detect ML task under the hood based on label column values. Currently we support:
- ModelTaskType.BINARY - ModelTaskType.MULTICLASS - ModelTaskType.REGRESSION

But for certain search datasets you can pass parameter to FeaturesEnricher with correct ML taks type: ```python from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey, ModelTaskType

enricher = FeaturesEnricher( searchkeys={"subscriptionactivationdate": SearchKey.DATE}, modeltask_type=ModelTaskType.REGRESSION ) ```

Time Series prediction support

Time series prediction supported as ModelTaskType.REGRESSION or ModelTaskType.BINARY tasks with time series specific cross-validation split: * Scikit-learn time series cross-validation - CVType.time_series parameter * Blocked time series cross-validation - CVType.blocked_time_series parameter

To initiate feature search you can pass cross-validation type parameter to FeaturesEnricher with time series specific CV type: ```python from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey, CVType

enricher = FeaturesEnricher( searchkeys={"salesdate": SearchKey.DATE}, cv=CVType.time_series ) ```

If you're working with multivariate time series, you should specify id columns of individual univariate series in FeaturesEnricher. For example, if you have a dataset predicting sales for different stores and products, you should specify store and product id columns as follows: python enricher = FeaturesEnricher( search_keys={ "sales_date": SearchKey.DATE, }, id_columns=["store_id", "product_id"], cv=CVType.time_series ) Pre-process search dataset in case of time series prediction:
sort rows in dataset according to observation order, in most cases - ascending order by date/datetime.

Accuracy and uplift metrics calculations

FeaturesEnricher automaticaly calculates model metrics and uplift from new relevant features either using calculate_metrics() method or calculate_metrics=True parameter in fit or fit_transform methods (example below).
You can use any model estimator with scikit-learn compartible interface, some examples are: * All Scikit-Learn supervised models * Xgboost * LightGBM * CatBoost

In addition to that list, you can define custom evaluation metric function using scikit-learn make_scorer, for example SMAPE.

By default, calculate_metrics() method calculates evaluation metric with the same cross-validation split as selected for FeaturesEnricher.fit() by parameter cv = CVType.<cross-validation-split>.
But you can easily define new split by passing child of BaseCrossValidator to parameter cv in calculate_metrics().

Example with more tips-and-tricks: ```python from upgini.features_enricher import FeaturesEnricher from upgini.metadata import SearchKey

enricher = FeaturesEnricher(searchkeys={"registrationdate": SearchKey.DATE})

Fit with default setup for metrics calculation

CatBoost will be used

enricher.fit(X, y, evalset=evalset, calculate_metrics=True)

LightGBM estimator for metrics

customestimator = LGBMRegressor() enricher.calculatemetrics(estimator=custom_estimator)

Custom metric function to scoring param (callable or name)

customscoring = "RMSLE" enricher.calculatemetrics(scoring=custom_scoring)

Custom cross validator

customcv = TimeSeriesSplit(nsplits=5) enricher.calculatemetrics(cv=customcv)

All this custom parameters could be combined in both methods: fit, fittransform and calculatemetrics:

enricher.fit(X, y, evalset, calculatemetrics=True, estimator=customestimator, scoring=customscoring, cv=custom_cv) ```

More tips-and-tricks

Automated feature generation from columns in a search dataset

If a training dataset has a text column, you can generate additional embeddings from it using instructed embeddings generation with LLMs and data augmentation from external sources, just like Upgini does for all records from connected data sources.

For most cases, this gives better results than direct embeddings generation from a text field. Currently, Upgini has two LLMs connected to a search engine - GPT-3.5 from OpenAI and GPT-J.

To use this feature, pass the column names as arguments to the generate_features parameter. You can use up to 2 columns.

Here's an example for generating features from the "description" and "summary" columns:

python enricher = FeaturesEnricher( search_keys={"date": SearchKey.DATE}, generate_features=["description", "summary"] )

With this code, Upgini will generate LLM embeddings from text columns and then check them for predictive power for your ML task.

Finally, Upgini will return a dataset enriched by only relevant components of LLM embeddings.

Find features only give accuracy gain to existing data in the ML model

If you already have features or other external data sources, you can specifically search new datasets & features only give accuracy gain "on top" of them.

Just leave all these existing features in the labeled training dataset and Upgini library automatically use them during feature search process and as a baseline ML model to calculate accuracy metric uplift. Only features which improve accuracy will return.

Check robustness of accuracy improvement from external features

You can validate external features robustness on out-of-time dataset using eval_set parameter: ```python

load train dataset

traindf = pd.readcsv("train.csv") trainidsandfeatures = traindf.drop(columns="label") trainlabel = traindf["label"]

load out-of-time validation dataset

evaldf = pd.readcsv("validation.csv") evalidsandfeatures = evaldf.drop(columns="label") evallabel = evaldf["label"]

create FeaturesEnricher

enricher = FeaturesEnricher(searchkeys={"registrationdate": SearchKey.DATE})

now we fit WITH eval_set parameter to calculate accuracy metrics on Out-of-time dataset.

the output will contain quality metrics for both the training data set and

the eval set (validation OOT data set)

enricher.fit( trainidsandfeatures, trainlabel, evalset = [(evalidsandfeatures, eval_label)] ) ```

Requirements for out-of-time dataset

• Same data schema as for search initialization X dataset
• Pandas dataframe representation

There are 3 options to pass out-of-time without labels: python enricher.fit( train_ids_and_features, train_label, eval_set = [ (eval_ids_and_features_1,), # Just tuple of 1 element (eval_ids_and_features_2, None), # None as labels (eval_ids_and_features_3, [np.nan] * len(eval_ids_and_features_3)), # List or Series of the same size as eval X ] )

Control feature stability with PSI parameters

FeaturesEnricher supports Population Stability Index (PSI) calculation on evalset to evaluate feature stability over time. You can control this behavior using stability parameters in fit and `fittransform` methods:

```python enricher = FeaturesEnricher( searchkeys={"registrationdate": SearchKey.DATE} )

Control feature stability during fit

enricher.fit( X, y, stabilitythreshold=0.2, # PSI threshold: features with PSI above this value will be dropped stabilityagg_func="max" # Aggregation function for stability values: "max", "min", "mean" )

Same parameters work for fit_transform

enricheddf = enricher.fittransform( X, y, stabilitythreshold=0.1, # Stricter threshold for more stable features stabilityagg_func="mean" # Use mean aggregation instead of max ) ```

Stability parameters: - stability_threshold (float, default=0.2): PSI threshold value. Features with PSI below this threshold will be excluded from the final feature set. Lower values mean stricter stability requirements. - stability_agg_func (str, default="max"): Function to aggregate PSI values across time intervals. Options: "max" (most conservative), "min" (least conservative), "mean" (balanced approach).

PSI (Population Stability Index) measures how much feature distribution changes over time. Lower PSI values indicate more stable features, which are generally more reliable for production ML models.

Use custom loss function in feature selection & metrics calculation

FeaturesEnricher can be initialized with additional string parameter loss. Depending on ML-task, you can use the following loss functions: - regression: regression, regression_l1, huber, poisson, quantile, mape, gamma, tweedie; - binary: binary; - multiclass: multiclass, multiclassova.

For instance, if your target variable has a Poisson distribution (count of events, number of customers in the shop and so on), you should try to use loss="poisson" to improve quality of feature selection and get better evaluation metrics.

Usage example: python enricher = FeaturesEnricher( search_keys={"date": SearchKey.DATE}, loss="poisson", model_task_type=ModelTaskType.REGRESSION ) enriched_dataframe.fit(X, y)

Exclude premium data sources from fit, transform and metrics calculation

fit, fit_transform, transform and calculate_metrics methods of FeaturesEnricher can be used with parameter exclude_features_sources that allows to exclude Trial or Paid features from Premium data sources: python enricher = FeaturesEnricher( search_keys={"subscription_activation_date": SearchKey.DATE} ) enricher.fit(X, y, calculate_metrics=False) trial_features = enricher.get_features_info()[enricher.get_features_info()["Feature type"] == "Trial"]["Feature name"].values.tolist() paid_features = enricher.get_features_info()[enricher.get_features_info()["Feature type"] == "Paid"]["Feature name"].values.tolist() enricher.calculate_metrics(exclude_features_sources=(trial_features + paid_features)) enricher.transform(X, exclude_features_sources=(trial_features + paid_features))

Turn off autodetection for search key columns

Upgini has autodetection of search keys on by default. To turn off use autodetect_search_keys=False:

```python enricher = FeaturesEnricher( searchkeys={"date": SearchKey.DATE}, autodetectsearch_keys=False, )

enricher.fit(X, y) ```

Turn off removing of target outliers

Upgini detect rows with target outlier for regression tasks. By default such rows are dropped on metrics calculation. To turn off removing of target outlier rows use parameter remove_outliers_calc_metrics=False in fit, fittransform or calculatemetrics methods:

```python enricher = FeaturesEnricher( search_keys={"date": SearchKey.DATE}, )

enricher.fit(X, y, removeoutlierscalc_metrics=False) ```

Turn off generating features on search keys

Upgini tries to generate features on email, date and datetime search keys. By default this generation is enabled. To disable it use parameter generate_search_key_features of FeaturesEnricher constructor:

python enricher = FeaturesEnricher( search_keys={"date": SearchKey.DATE}, generate_search_key_features=False, )

Open up all capabilities of Upgini

Register and get a free API key for exclusive data sources and features: 600 mln+ phone numbers, 350 mln+ emails, 2^32 IP addresses

|Benefit|No Sign-up | Registered user | |--|--|--| |Enrichment with date/datetime, postal/ZIP code and country keys | Yes | Yes | |Enrichment with phone number, hashed email/HEM and IP-address keys | No | Yes | |Email notification on search task completion | No | Yes | |Automated feature generation with LLMs from columns in a search dataset| Yes, till 12/05/23 | Yes | |Email notification on new data source activation | No | Yes |

How to share data/features with a community ?

You may publish ANY data which you consider as royalty / license free (Open Data) and potentially valuable for ML applications for community usage:
1. Please Sign Up here 2. Copy Upgini API key from profile and upload your data from Upgini python library with this key: ```python import pandas as pd from upgini.metadata import SearchKey from upgini.ads import uploaduserads import os os.environ["UPGINIAPIKEY"] = "yourlongstringapikeygoeshere"

you can define custom search key which might not be supported yet, just use SearchKey.CUSTOM_KEY type

sampledf = pd.readcsv("pathtodatasamplefile") uploaduserads("test", sampledf, { "city": SearchKey.CUSTOMKEY, "stats_date": SearchKey.DATE }) ``` 3. After data verification, search results on community data will be available usual way.

Getting Help & Community

Please note, that we are still in a beta stage. Requests and support, in preferred order

Please try to create bug reports that are: - reproducible - include steps to reproduce the problem. - specific - include as much detail as possible: which Python version, what environment, etc. - unique - do not duplicate existing opened issues. - scoped to a Single Bug - one bug per report.

Contributing

We are not a large team, so we probably won't be able to: - implement smooth integration with most common low-code ML libraries and platforms (PyCaret, H2O AutoML, etc. ) - implement all possible data verification and normalization capabilities for different types of search keys And we need some help from the community!

So, we'll be happy about every pull request you open and issue you find to make this library more incredible. Please note that it might sometimes take us a while to get back to you. For major changes, please open an issue first to discuss what you would like to change

Developing

Some convenient ways to start contributing are:
Open in Visual Studio Code You can remotely open this repo in VS Code without cloning or automatically clone and open it inside a docker container.
Gitpod You can use Gitpod to launch a fully functional development environment right in your browser.

Useful links

• Simple sales predictions as a template notebook
• Full list of Kaggle Guides & Examples
• Project on PyPI
• More perks for registered users

<sup>Found mistype or a bug in code snippet? Our bad! Please report it here</sup>