sudo apt install -y postgresml-14

Ubuntu 22.04 ships with PostgreSQL 14, but if you have a different version installed on your system, just change`14` in the package name to your Postgres version. We currently support all versions supported by the community: Postgres 12 through16.

Ubuntu 22.04 ships with PostgreSQL 14, but if you have a different version installed on your system, just change`14` in the package name to your Postgres version. We currently support all versions supported by the community: Postgres 12 through15.

`pgrx` is open source and available from crates.io. We are currently using the`0.11.0` version. It's important that your`pgrx` version matches what we're using, since there are some hard dependencies between our code and`pgrx`.

`pgrx` is open source and available from crates.io. We are currently using the`0.10.0` version. It's important that your`pgrx` version matches what we're using, since there are some hard dependencies between our code and`pgrx`.

To install`pgrx`, simply run:

Before using`pgrx`, it needs to be initialized against the installed version of PostgreSQL. In this example, we'll be using the Ubuntu 22.04 default PostgreSQL 14 installation:

virtualenv pgml-venv&& \

source pgml-venv/bin/activate&& \

pip install -r requirements.txt&& \

pip install -r requirements-autogptq.txt&& \

pip install -r requirements-xformers.txt --no-dependencies

1. Go to[https://postgresml.org/signup](https://postgresml.org/signup)

2. Sign up using your email or using Google or Github authentication

3. Login using your account

4.[data-pre-processing.md](../machine-learning/supervised-learning/data-pre-processing.md"mention")

There are 3 steps to preprocessing data:

*[Encoding](#categorical-encodings) categorical values into quantitative values

*[Imputing](#imputing-missing-values) NULL values to some quantitative value

*[Scaling](#scaling-values) quantitative values across all variables to similar ranges

*[Encoding](../../../../../pgml-dashboard/content/docs/guides/training/preprocessing.md#categorical-encodings) categorical values into quantitative values

*[Imputing](../../../../../pgml-dashboard/content/docs/guides/training/preprocessing.md#imputing-missing-values) NULL values to some quantitative value

*[Scaling](../../../../../pgml-dashboard/content/docs/guides/training/preprocessing.md#scaling-values) quantitative values across all variables to similar ranges

These preprocessing steps may be specified on a per-column basis to the[train()](../../../../../docs/guides/training/overview) function. By default, PostgresML does minimal preprocessing on training data, and will raise an error during analysis if NULL values are encountered without a preprocessor. All types other than`TEXT` are treated as quantitative variables and cast to floating point representations before passing them to the underlying algorithm implementations.

|**name**|**description**|

| ---------| -----------------------------------------------------------------------------------------------------------------------------------------------|

|`none`|**Default** - Casts the variable to a 32-bit floating point representation compatible with numerics. This is the default for non-`TEXT` values.|

|`target`| Encodes the variable as theaverage value of the target label for all members of the category. This is the default for`TEXT` variables.|

|`target`| Encodes the variable as themean value of the target label for all members of the category. This is the default for`TEXT` variables.|

|`one_hot`| Encodes the variable as multiple independent boolean columns.|

|`ordinal`| Encodes the variable as integer values provided by their position in the input array. NULLS are always 0.|

{% tab title="JavaScript" %}

model=pgml.newModel()

constmodel=pgml.newModel()

{% endtab %}

{% endtabs %}

{% tab title="JavaScript" %}

model=pgml.newModel(

name="hkunlp/instructor-base",

parameters={instruction:"Represent the Wikipedia document for retrieval:"}

constmodel=pgml.newModel(

"hkunlp/instructor-base",

"pgml",

{ instruction:"Represent the Wikipedia document for retrieval:" }

)

{% endtab %}

{% tab title="JavaScript" %}

model=pgml.newModel(name="text-embedding-ada-002",source="openai")

constmodel=pgml.newModel("text-embedding-ada-002","openai")

{% endtab %}

{% endtabs %}

{% tab title="JavaScript" %}

splitter=pgml.newSplitter()

constsplitter=pgml.newSplitter()

{% endtab %}

{% endtabs %}

{% tab title="JavaScript" %}

splitter=pgml.newSplitter(

name="recursive_character",

parameters={chunk_size:1500, chunk_overlap:40}

"recursive_character",

{chunk_size:1500, chunk_overlap:40}

)

{% endtab %}

{% tab title="JavaScript" %}

model=pgml.newModel()

splitter=pgml.newSplitter()

pipeline=pgml.newPipeline("test_pipeline", model, splitter)

constmodel=pgml.newModel()

constsplitter=pgml.newSplitter()

constpipeline=pgml.newPipeline("test_pipeline", model, splitter)

awaitcollection.add_pipeline(pipeline)

{% endtab %}

{% tab title="JavaScript" %}

model=pgml.newModel()

splitter=pgml.newSplitter()

pipeline=pgml.newPipeline("test_pipeline", model, splitter, {

"full_text_search": {

active: True,

configuration:"english"

constmodel=pgml.newModel()

constsplitter=pgml.newSplitter()

constpipeline=pgml.newPipeline("test_pipeline", model, splitter, {

full_text_search: {

active:true,

configuration:"english"

}

})

awaitcollection.add_pipeline(pipeline)

{% endtab %}

{% endtabs %}

By default the SDK uses HNSW indexes to efficiently perform vector recall. The default HNSW index sets`m` to 16 and`ef_construction` to 64. These defaults can be customized when the Pipeline is created.

{% tabs %}

{% tab title="Python" %}

model= Model()

splitter= Splitter()

pipeline= Pipeline("test_pipeline", model, splitter, {

"hnsw": {

"m":16,

"ef_construction":64

}

})

await collection.add_pipeline(pipeline)

{% endtab %}

{% tab title="JavaScript" %}

constmodel=pgml.newModel()

constsplitter=pgml.newSplitter()

constpipeline=pgml.newPipeline("test_pipeline", model, splitter, {

hnsw: {

m:16,

ef_construction:64

}

})

awaitcollection.add_pipeline(pipeline)

{% endtab %}

{% endtabs %}

{% tab title="JavaScript" %}

pipeline=pgml.newPipeline("test_pipeline")

collection=pgml.newCollection("test_collection")

results=awaitcollection.query().vector_recall("Why is PostgresML the best?", pipeline).fetch_all()

constpipeline=pgml.newPipeline("test_pipeline")

constcollection=pgml.newCollection("test_collection")

constresults=awaitcollection.query().vector_recall("Why is PostgresML the best?", pipeline).fetch_all()

{% endtab %}

{% endtabs %}

Pipelines can be disabled or removed to prevent them from running automatically when documents are upserted.

{% tabs %}

{% tab title="Python" %}

{% tab title="JavaScript" %}

pipeline=pgml.newPipeline("test_pipeline")

collection=pgml.newCollection("test_collection")

constpipeline=pgml.newPipeline("test_pipeline")

constcollection=pgml.newCollection("test_collection")

awaitcollection.disable_pipeline(pipeline)

{% endtab %}

Disabled pipelines can be re-enabled.

{% tabs %}

{% tab title="Python" %}

{% tab title="JavaScript" %}

pipeline=pgml.newPipeline("test_pipeline")

collection=pgml.newCollection("test_collection")

constpipeline=pgml.newPipeline("test_pipeline")

constcollection=pgml.newCollection("test_collection")

awaitcollection.enable_pipeline(pipeline)

{% endtab %}

{% endtab %}

{% tab title="JavaScript" %}

pipeline=pgml.newPipeline("test_pipeline")

collection=pgml.newCollection("test_collection")

awaitcollection.remove_pipeline(pipeline)

<preclass="language-javascript"><codeclass="lang-javascript">const pipeline = pgml.newPipeline("test_pipeline")

<strong>const collection = pgml.newCollection("test_collection")

</strong>await collection.remove_pipeline(pipeline)

</code></pre>

{% endtab %}

{% endtabs %}