merge

mckinsey · Jul 6, 2024 · 1600b55 · 1600b55
1 parent 5da09eb
commit 1600b55
Show file tree

Hide file tree

Showing 9 changed files with 272 additions and 105 deletions.
diff --git a/vizro-ai/changelog.d/20240626_224646_anna_xiong_azure_openai.md b/vizro-ai/changelog.d/20240626_224646_anna_xiong_azure_openai.md
@@ -0,0 +1,48 @@
+<!--
+A new scriv changelog fragment.
+
+Uncomment the section that is right (remove the HTML comment wrapper).
+-->
+
+<!--
+### Highlights ✨
+
+- A bullet item for the Highlights ✨ category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Removed
+
+- A bullet item for the Removed category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Added
+
+- A bullet item for the Added category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Changed
+
+- A bullet item for the Changed category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Deprecated
+
+- A bullet item for the Deprecated category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Fixed
+
+- A bullet item for the Fixed category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Security
+
+- A bullet item for the Security category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
diff --git a/vizro-ai/src/vizro_ai/chains/_llm_models.py b/vizro-ai/src/vizro_ai/chains/_llm_models.py
@@ -4,63 +4,34 @@
 from langchain_core.language_models.chat_models import BaseChatModel
 from langchain_openai import ChatOpenAI
 
-# TODO is there a better way to handle this import?
 try:
     from langchain_anthropic import ChatAnthropic
 except ImportError:
     ChatAnthropic = None
 
-# TODO to be removed, just use BaseChatModel should be enough
-LLM_MODELS = Union[ChatOpenAI]
-
-# TODO constant of model inventory, can be converted to yaml and link to docs
-PREDEFINED_MODELS: Dict[str, Dict[str, Union[int, BaseChatModel]]] = {
-    "gpt-3.5-turbo-0613": {
-        "max_tokens": 4096,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-4-0613": {
-        "max_tokens": 8192,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-3.5-turbo-1106": {
-        "max_tokens": 16385,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-4-1106-preview": {
-        "max_tokens": 128000,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-3.5-turbo-0125": {
-        "max_tokens": 16385,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-3.5-turbo": {
-        "max_tokens": 16385,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-4-turbo": {
-        "max_tokens": 128000,
-        "wrapper": ChatOpenAI,
-    },
-    "gpt-4o": {
-        "max_tokens": 128000,
-        "wrapper": ChatOpenAI,
-    },
-}
-
-# TODO add new wrappers in if new model support is added
-if ChatAnthropic is not None:
-    PREDEFINED_MODELS = {
-        **PREDEFINED_MODELS,
-        **{"claude-3-haiku-20240307": {"max_tokens": 200000, "wrapper": ChatAnthropic}},
-        **{"claude-3-sonnet-20240229": {"max_tokens": 200000, "wrapper": ChatAnthropic}},
-    }
 
+SUPPORTED_MODELS = {
+    "OpenAI": [
+        "gpt-4-0613",
+        "gpt-3.5-turbo-1106",
+        "gpt-4-1106-preview",
+        "gpt-3.5-turbo-0125",
+        "gpt-3.5-turbo",
+        "gpt-4-turbo",
+        "gpt-4o",
+    ],
+    "Anthropic": [
+        "claude-3-haiku-20240307",
+        "claude-3-sonnet-20240229",
+    ],
+}
 
+DEFAULT_WRAPPER_MAP: Dict[str, BaseChatModel] = {"OpenAI": ChatOpenAI, "Anthropic": ChatAnthropic}
 DEFAULT_MODEL = "gpt-3.5-turbo"
 DEFAULT_TEMPERATURE = 0
 
+model_to_vendor = {model: key for key, models in SUPPORTED_MODELS.items() for model in models}
+
 
 def _get_llm_model(model: Optional[Union[ChatOpenAI, str]] = None) -> BaseChatModel:
     """Fetches and initializes an instance of the LLM.
@@ -77,10 +48,17 @@ def _get_llm_model(model: Optional[Union[ChatOpenAI, str]] = None) -> BaseChatMo
     """
     if not model:
         return ChatOpenAI(model_name=DEFAULT_MODEL, temperature=DEFAULT_TEMPERATURE)
-    if isinstance(model, ChatOpenAI):
+
+    if isinstance(model, BaseChatModel):
         return model
-    if isinstance(model, str) and model in PREDEFINED_MODELS:
-        return PREDEFINED_MODELS.get(model)["wrapper"](model_name=model, temperature=DEFAULT_TEMPERATURE)
+
+    if isinstance(model, str):
+        if any(model in model_list for model_list in SUPPORTED_MODELS.values()):
+            vendor = model_to_vendor[model]
+            if DEFAULT_WRAPPER_MAP.get(vendor) is None:
+                raise ValueError(f"Addtitional library to support {vendor} models is not installed.")
+            return DEFAULT_WRAPPER_MAP.get(vendor)(model_name=model, temperature=DEFAULT_TEMPERATURE)
+
     raise ValueError(
         f"Model {model} not found! List of available model can be found at https://vizro.readthedocs.io/projects/vizro-ai/en/latest/pages/explanation/faq/#which-llms-are-supported-by-vizro-ai"
     )
@@ -100,4 +78,6 @@ def _get_model_name(model):
 
 
 if __name__ == "__main__":
-    llm_chat_openai = _get_llm_model()
+    llm_chat_openai = _get_llm_model(model="gpt-3.5-turbo")
+    print(repr(llm_chat_openai))  # noqa: T201
+    print(llm_chat_openai.model_name)  # noqa: T201
diff --git a/vizro-ai/src/vizro_ai/dashboard/nodes/plan.py b/vizro-ai/src/vizro_ai/dashboard/nodes/plan.py
@@ -59,7 +59,7 @@ def create(self, model, df_metadata) -> Union[ComponentType, None]:
             return vm.AgGrid(id=self.component_id, figure=dash_ag_grid(data_frame=self.data_frame))
         elif self.component_type == "Card":
             return _get_proxy_model(
-                query=self.component_description, model=model, result_model=vm.Card, df_metadata=df_metadata
+                query=self.component_description, llm_model=model, result_model=vm.Card, df_metadata=df_metadata
             )
 
 
@@ -142,7 +142,7 @@ def create(self, model, available_components, df_metadata):
                 df_cols=_df_cols, df_sample=_df_sample, available_components=available_components
             )
             proxy = _get_proxy_model(
-                query=filter_prompt, model=model, result_model=result_proxy, df_metadata=df_metadata
+                query=filter_prompt, llm_model=model, result_model=result_proxy, df_metadata=df_metadata
             )
             logger.info(
                 f"`Control` proxy: {proxy.dict()}"
@@ -199,7 +199,7 @@ def create(self, model, df_metadata) -> Union[vm.Layout, None]:
 
         try:
             proxy = _get_proxy_model(
-                query=self.layout_description, model=model, result_model=LayoutProxyModel, df_metadata=df_metadata
+                query=self.layout_description, llm_model=model, result_model=LayoutProxyModel, df_metadata=df_metadata
             )
             actual = vm.Layout.parse_obj(proxy.dict(exclude={}))
         except (ValidationError, AttributeError) as e:
@@ -238,7 +238,7 @@ def _get_dashboard_plan(
     model: Union[ChatOpenAI],
     df_metadata: Dict[str, Dict[str, str]],
 ) -> DashboardPlanner:
-    return _get_proxy_model(query=query, model=model, result_model=DashboardPlanner, df_metadata=df_metadata)
+    return _get_proxy_model(query=query, llm_model=model, result_model=DashboardPlanner, df_metadata=df_metadata)
 
 
 def _print_dashboard_plan(dashboard_plan) -> None:

diff --git a/vizro-core/changelog.d/20240703_091955_huong_li_nguyen_fix_selector_return_type.md b/vizro-core/changelog.d/20240703_091955_huong_li_nguyen_fix_selector_return_type.md
@@ -0,0 +1,47 @@
+<!--
+A new scriv changelog fragment.
+
+Uncomment the section that is right (remove the HTML comment wrapper).
+-->
+
+<!--
+### Highlights ✨
+
+- A bullet item for the Highlights ✨ category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Removed
+
+- A bullet item for the Removed category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Added
+
+- A bullet item for the Added category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Changed
+
+- A bullet item for the Changed category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+<!--
+### Deprecated
+
+- A bullet item for the Deprecated category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
+
+### Fixed
+
+- Ensure that categorical selectors always return a list of values. ([#562](https://github.com/mckinsey/vizro/pull/562))
+
+<!--
+### Security
+
+- A bullet item for the Security category with a link to the relevant PR at the end of your entry, e.g. Enable feature XXX ([#1](https://github.com/mckinsey/vizro/pull/1))
+
+-->
diff --git a/vizro-core/examples/_dev/app.py b/vizro-core/examples/_dev/app.py
@@ -1,64 +1,49 @@
 """Dev app to try things out."""
 
-from typing import Optional
-
-import dash_bootstrap_components as dbc
 import pandas as pd
 import vizro.models as vm
 import vizro.plotly.express as px
-from dash import html
 from vizro import Vizro
-from vizro.figures import kpi_card
 from vizro.models.types import capture
 
-tips = px.data.tips
+df_stocks = px.data.stocks(datetimes=True)
+df_stocks_long = pd.melt(
+    df_stocks,
+    id_vars="date",
+    value_vars=["GOOG", "AAPL", "AMZN", "FB", "NFLX", "MSFT"],
+    var_name="stocks",
+    value_name="value",
+)
 
 
-@capture("figure")  # (1)!
-def custom_kpi_card(  # noqa: PLR0913
-    data_frame: pd.DataFrame,
-    value_column: str,
-    *,
-    value_format: str = "{value}",
-    agg_func: str = "sum",
-    title: Optional[str] = None,
-    icon: Optional[str] = None,
-) -> dbc.Card:  # (2)!
-    """Creates a custom KPI card."""
-    title = title or f"{agg_func} {value_column}".title()
-    value = data_frame[value_column].agg(agg_func)
+@capture("graph")
+def vizro_plot(data_frame, stocks_selected, **kwargs):
+    """Custom chart function."""
+    return px.line(data_frame[data_frame["stocks"].isin(stocks_selected)], **kwargs)
 
-    header = dbc.CardHeader(
-        [
-            html.H2(title),
-            html.P(icon, className="material-symbols-outlined") if icon else None,  # (3)!
-        ]
-    )
-    body = dbc.CardBody([value_format.format(value=value)])
-    return dbc.Card([header, body], className="card-kpi")
 
+df_stocks_long["value"] = df_stocks_long["value"].round(3)
 
 page = vm.Page(
-    title="Create your own KPI card",
-    layout=vm.Layout(grid=[[0, 1, -1, -1]] + [[-1, -1, -1, -1]] * 3),  # (4)!
+    title="My first page",
     components=[
-        vm.Figure(
-            figure=kpi_card(  # (5)!
-                data_frame=tips,
-                value_column="tip",
-                value_format="${value:.2f}",
-                icon="shopping_cart",
-                title="Default KPI card",
-            )
+        vm.Graph(
+            id="my_graph",
+            figure=vizro_plot(
+                data_frame=df_stocks_long,
+                stocks_selected=list(df_stocks_long["stocks"].unique()),
+                x="date",
+                y="value",
+                color="stocks",
+            ),
         ),
-        vm.Figure(
-            figure=custom_kpi_card(  # (6)!
-                data_frame=tips,
-                value_column="tip",
-                value_format="${value:.2f}",
-                icon="payment",
-                title="Custom KPI card",
-            )
+    ],
+    controls=[
+        vm.Parameter(
+            targets=["my_graph.stocks_selected"],
+            selector=vm.Dropdown(
+                options=[{"label": s, "value": s} for s in df_stocks_long["stocks"].unique()],
+            ),
         ),
     ],
 )

diff --git a/vizro-core/src/vizro/actions/_actions_utils.py b/vizro-core/src/vizro/actions/_actions_utils.py
@@ -138,12 +138,19 @@ def _get_parametrized_config(target: ModelID, ctd_parameters: List[CallbackTrigg
     config["data_frame"] = {}
 
     for ctd in ctd_parameters:
-        selector_value = ctd[
-            "value"
-        ]  # TODO: needs to be refactored so that it is independent of implementation details
+        # TODO: needs to be refactored so that it is independent of implementation details
+        selector_value = ctd["value"]
+
         if hasattr(selector_value, "__iter__") and ALL_OPTION in selector_value:  # type: ignore[operator]
             selector: SelectorType = model_manager[ctd["id"]]
-            selector_value = selector.options
+
+            # Even if options are provided as List[Dict], the Dash component only returns a List of values.
+            # So we need to ensure that we always return a List only as well to provide consistent types.
+            if all(isinstance(option, dict) for option in selector.options):
+                selector_value = [option["value"] for option in selector.options]
+            else:
+                selector_value = selector.options
+
         selector_value = _validate_selector_value_none(selector_value)
         selector_actions = _get_component_actions(model_manager[ctd["id"]])
 

diff --git a/vizro-core/src/vizro/models/_action/_action.py b/vizro-core/src/vizro/models/_action/_action.py
@@ -105,7 +105,7 @@ def _action_callback_function(
     ) -> Any:
         logger.debug("===== Running action with id %s, function %s =====", self.id, self.function._function.__name__)
         if logger.isEnabledFor(logging.DEBUG):
-            logger.debug("Action inputs:\n%s", pformat(inputs, depth=2, width=200))
+            logger.debug("Action inputs:\n%s", pformat(inputs, depth=3, width=200))
             logger.debug("Action outputs:\n%s", pformat(outputs, width=200))
 
         if isinstance(inputs, Mapping):