time-split 0.5.0__tar.gz → 0.6.0__tar.gz

{time_split-0.5.0 → time_split-0.6.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: time-split
-Version: 0.5.0
+Version: 0.6.0
 Summary: Time-based k-fold validation splits for heterogeneous data.
 Home-page: https://github.com/rsundqvist/time-split
 Author: Richard Sundqvist
@@ -19,13 +19,10 @@ Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3 :: Only
 Provides-Extra: plotting
-Provides-Extra: streamlit
-Requires-Dist: croniter ; extra == "streamlit"
-Requires-Dist: matplotlib ; extra == "plotting" or extra == "streamlit"
+Requires-Dist: matplotlib ; extra == "plotting"
 Requires-Dist: pandas (>=2.0.3)
 Requires-Dist: rics (>=4.1.1)
-Requires-Dist: seaborn ; extra == "plotting" or extra == "streamlit"
-Requires-Dist: streamlit ; extra == "streamlit"
+Requires-Dist: seaborn ; extra == "plotting"
 Project-URL: Bug Tracker, https://github.com/rsundqvist/time-split/issues
 Project-URL: Changelog, https://github.com/rsundqvist/time-split/blob/master/CHANGELOG.md
 Project-URL: Documentation, https://time-split.readthedocs.io
@@ -73,7 +70,7 @@ strong non-stationary characteristics.
 
 ## Experimenting with parameters
 The **Time Fold Explorer** application
-(available [here](http://20.240.208.202/time-split/?data=1554942900-1557610200&schedule=0+0+%2A+%2A+MON%2CFRI&n_splits=2&step=2&show_removed=True))
+(available [here](https://time-split.streamlit.app/?data=1554942900-1557610200&schedule=0+0+%2A+%2A+MON%2CFRI&n_splits=2&step=2&show_removed=True))
 is designed to help evaluate the effects of different
 [parameters](https://time-split.readthedocs.io/en/stable/#parameter-overview).
 To start it locally using

{time_split-0.5.0 → time_split-0.6.0}/README.md

@@ -39,7 +39,7 @@ strong non-stationary characteristics.
 
 ## Experimenting with parameters
 The **Time Fold Explorer** application
-(available [here](http://20.240.208.202/time-split/?data=1554942900-1557610200&schedule=0+0+%2A+%2A+MON%2CFRI&n_splits=2&step=2&show_removed=True))
+(available [here](https://time-split.streamlit.app/?data=1554942900-1557610200&schedule=0+0+%2A+%2A+MON%2CFRI&n_splits=2&step=2&show_removed=True))
 is designed to help evaluate the effects of different
 [parameters](https://time-split.readthedocs.io/en/stable/#parameter-overview).
 To start it locally using

{time_split-0.5.0 → time_split-0.6.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "time-split"
-version = "0.5.0"
+version = "0.6.0"
 description = "Time-based k-fold validation splits for heterogeneous data."
 authors = ["Richard Sundqvist <richard.sundqvist@live.se>"]
 
@@ -40,13 +40,8 @@ rics = ">=4.1.1"
 matplotlib = { version = "*", optional = true }
 seaborn = { version = "*", optional = true }
 
-# explorer
-streamlit = { version = "*", optional = true }
-croniter = { version = "*", optional = true }
-
 [tool.poetry.extras]
 plotting = ["matplotlib", "seaborn"]
-streamlit = ["matplotlib", "seaborn", "streamlit", "croniter"]
 
 [tool.poetry.group.manual-extras.dependencies]
 # Extras that must be available for docs + tests go here. Doc/test-only deps go in their respective groups.

{time_split-0.5.0 → time_split-0.6.0}/src/time_split/__init__.py

@@ -11,6 +11,6 @@ __all__ = [
     "__version__",  # Make MyPy happy
 ]
 
-__version__ = "0.5.0"
+__version__ = "0.6.0"
 
 _logging.getLogger(__name__).addHandler(_logging.NullHandler())

{time_split-0.5.0 → time_split-0.6.0}/src/time_split/_docstrings.py

@@ -17,7 +17,7 @@ def docs(__func: Callable[P, T], /) -> Callable[P, T]:
     __func.__doc__ = __func.__doc__.format(**_DOCSTRINGS)
 
     if isinstance(__func, type):
-        __func.__init__ = deprecated_params(__func.__init__)
+        __func.__init__ = deprecated_params(__func.__init__)  # type: ignore[misc]
     else:
         __func = deprecated_params(__func)
 

time_split-0.5.0/src/time_split/streamlit/README.md

@@ -1,82 +0,0 @@
-# Time Split  <!-- omit in toc -->
-Time-based k-fold validation splits for heterogeneous data.
-
------------------
-[![PyPI - Version](https://img.shields.io/pypi/v/time-split.svg)](https://pypi.python.org/pypi/time-split)
-[![PyPI - Python Version](https://img.shields.io/pypi/pyversions/time-split.svg)](https://pypi.python.org/pypi/time-split)
-[![Tests](https://github.com/rsundqvist/time-split/workflows/tests/badge.svg)](https://github.com/rsundqvist/time-split/actions?workflow=tests)
-[![Codecov](https://codecov.io/gh/rsundqvist/time-split/branch/master/graph/badge.svg)](https://codecov.io/gh/rsundqvist/time-split)
-[![Read the Docs](https://readthedocs.org/projects/time-split/badge/)](https://time-split.readthedocs.io/)
-[![PyPI - License](https://img.shields.io/pypi/l/time-split.svg)](https://pypi.python.org/pypi/time-split)
-[![Docker Image Size (tag)](https://img.shields.io/docker/image-size/rsundqvist/time-split/latest?logo=docker&label=time-split)](https://hub.docker.com/r/rsundqvist/time-split/)
-
-<div align="center">
-  <img alt="Plotted folds on a two-by-two grid." 
-       title="Examples" height="300" width="1200" 
-  src="https://raw.githubusercontent.com/rsundqvist/time-split/master/docs/2x2-examples.jpg"><br>
-</div>
-
-Folds plotted on a two-by-two grid. See the
-[examples](https://time-split.readthedocs.io/en/stable/auto_examples/index.html) page for more.
-
-# About this image
-The **Time Fold Explorer** application
-(available [here](http://20.240.208.202/time-split/?data=1554942900-1557610200&schedule=0+0+%2A+%2A+MON%2CFRI&n_splits=2&step=2&show_removed=True))
-is designed to help evaluate the effects of different
-[parameters](https://time-split.readthedocs.io/en/stable/#parameter-overview).
-To start it locally using
-[![Docker Image Size (tag)](https://img.shields.io/docker/image-size/rsundqvist/time-split/latest?logo=docker&label=time-split)](https://hub.docker.com/r/rsundqvist/time-split/)
-Docker, run
-```sh
-docker run -p 8501:8501 rsundqvist/time-split
-```
-in the terminal. You may use
-[`create_explorer_link()`](https://time-split.readthedocs.io/en/stable/generated/time_split.support.html#time_split.support.create_explorer_link)
-to build application URLs with preselected splitting parameters.
-
-# Custom datasets
-To bundle datasets, mount a configuration file (determined by 
-[`DATASETS_CONFIG_PATH='/home/streamlit/datasets.toml'`](https://time-split.readthedocs.io/en/stable/generated/time_split.streamlit.config.html#time_split.streamlit.config.DATASETS_CONFIG_PATH)
-). The `DatasetConfig` struct has the following keys:
-
-* `label`: Name shown in the UI. Defaults to section header (i.e. *"my-dataset"* below).
-* `path`: Required. First argument to the `pandas` read function.
-* `index`: Required. Must be datetime-like.
-* `aggregations`: Optional. Determines function to use in the `📈 Aggregations per fold` tab.
-* `description`: Optional. The first line will be used as the summary in the UI.
-* `read_function_kwargs`: Keyword arguments for the `pandas` read function used (as determined by the `path`).
-
-Additional dependencies are required for remote filesystems. You may use `EXTRA_PIP_PACKAGES=s3fs` to install
-dependencies for the S3 paths used below.
-
-```toml
-[my-dataset]
-label = "Dataset name"
-path = "s3://my-bucket/data/title_basics.csv"
-index = "from"
-aggregations = { runtimeMinutes = "min", isAdult = "mean" }
-description = """This is the summary.
-
-Simplified version of the [Title basics](https://developer.imdb.com/non-commercial-datasets/#titlebasicstsvgz) IMDB
-dataset. The description supports Markdown syntax.
-"""
-[my-dataset.read_function_kwargs]
-# Valid options depend on the read function used.
-```
-
-Multiple datasets may be configured in their own top-level sections. Labels must be unique.
-
-# Environment variables
-See [time_split.streamlit.config](https://time-split.readthedocs.io/en/stable/generated/time_split.streamlit.config.html) 
-for configurable values. Use `true|false` for boolean variables. Documentation for the underlying framework (Streamlit)
-is available 
-[here](https://docs.streamlit.io/develop/concepts/configuration/options/).
-
-## User choice
-Users may *lower* some configured values by using the Performance tweaker widget in the `❔ About tab` of application. To 
-set a lower default, add a `DEFAULT_`-prefix to the regular name.
-```bash
-PLOT_AGGREGATIONS_PER_FOLD=true
-DEFAULT_PLOT_AGGREGATIONS_PER_FOLD=false
-```
-This will disable the (expensive) per-column fold aggregation figures, but users who need them can turn them back on.

time_split-0.5.0/src/time_split/streamlit/__init__.py

@@ -1 +0,0 @@
-"""Reusable components used by the **Time Fold Explorer** application."""

time_split-0.5.0/src/time_split/streamlit/_logging.py

@@ -1,75 +0,0 @@
-import logging
-from typing import Any
-
-import pandas as pd
-import streamlit as st
-from rics.collections.dicts import flatten_dict
-from streamlit.runtime.scriptrunner import get_script_run_ctx
-
-from time_split.streamlit import config
-from time_split.streamlit.widgets.types import QueryParams
-
-LOGGER = logging.getLogger("time_split.streamlit")
-
-
-def log_perf(
-    message: str,
-    df: pd.DataFrame | dict[str, pd.DataFrame],
-    seconds: float,
-    *,
-    extra: dict[str, Any],
-    level: int = logging.INFO,
-) -> str:
-    if isinstance(df, dict):
-        frames: dict[str, Any] = {key: _get_frame_extras(df) for key, df in df.items()}
-        frames["__sum__"] = pd.DataFrame.from_dict(frames, orient="index").sum().to_dict()
-        frames["n_frames"] = len(df)
-        frame_extras = flatten_dict(frames)
-    else:
-        frame_extras = _get_frame_extras(df)
-
-    remote_ip, session_id = get_session_data()
-    extra = extra | {
-        "session_id": session_id,
-        "remote_ip": remote_ip,
-        "duration_ms": int(1000 * seconds),
-        **frame_extras,
-        **config.get_values(),
-        **QueryParams.get().to_dict(prefix="query."),
-    }
-    message = message.format_map(extra)
-    LOGGER.log(level, message.replace("`", "") + f" | {extra=}", extra=extra)
-    return message
-
-
-def _get_frame_extras(df: pd.DataFrame) -> dict[str, int]:
-    return {"size": df.size, "rows": len(df), "columns": len(df.columns)}
-
-
-class StreamlitLoggingHandler(logging.Handler):
-    def emit(self, record: logging.LogRecord) -> None:
-        message = record.getMessage()
-
-        if record.levelno >= logging.ERROR:
-            st.error(message, icon="🚨")
-        elif record.levelno >= logging.WARNING:
-            st.warning(message, icon="⚠️")
-        elif record.levelno >= logging.INFO:
-            st.info(message, icon="ℹ️")
-        else:
-            st.code(record.levelname + ": " + record.getMessage())
-
-
-def get_session_data() -> tuple[str, str]:
-    ctx = get_script_run_ctx(suppress_warning=True)
-    if ctx is None:
-        return "", ""
-
-    return _get_session_data(ctx.session_id), ctx.session_id
-
-
-def _get_session_data(session_id: str) -> str:
-    from streamlit.runtime import get_instance
-
-    client = get_instance().get_client(session_id)
-    return client.request.remote_ip  # type: ignore[no-any-return, union-attr]

time_split-0.5.0/src/time_split/streamlit/_support.py

@@ -1,67 +0,0 @@
-from pprint import pformat
-from typing import Never
-
-import pandas as pd
-import streamlit as st
-
-from time_split.streamlit import config
-from time_split.streamlit.widgets.display._code import CodeWidget
-from time_split.types import DatetimeIndexSplitterKwargs
-
-
-def enforce_max_splits(
-    all_splits: int, kwargs: DatetimeIndexSplitterKwargs, limits: tuple[pd.Timestamp, pd.Timestamp]
-) -> None:
-    """Check the number of splits against the configured limit, rising if exceeded."""
-    if all_splits > config.MAX_SPLITS:
-        _too_many_splits(all_splits, kwargs, limits)
-
-
-@st.experimental_dialog("🚨 Too many splits")
-def _too_many_splits(
-    all_splits: int,
-    kwargs: DatetimeIndexSplitterKwargs,
-    limits: tuple[pd.Timestamp, pd.Timestamp],
-) -> Never:
-    st.error(f"Maximum number of splits ({config.MAX_SPLITS}) exceeded.", icon="🚨")
-    st.write("Number of splits is restricted for performance reasons.")
-    strings = CodeWidget("string").convert({**kwargs, "available": limits})
-    st.code("kwargs = " + pformat(strings, indent=0, width=40, depth=2))
-    st.write(
-        f"The arguments above produces `len(splits)={all_splits} > {config.MAX_SPLITS}=MAX_SPLITS`."
-        " Try using different parameters to reduce the number of folds."
-    )
-    st.stop()
-
-
-def get_about() -> str:
-    from time_split import __version__
-
-    return f"""## Time Fold Explorer
-This applications is designed to help experiment with `time-split` splitting parameters.
-
-See https://time-split.readthedocs.io/ for library documentation.
-
-#### 🚀 Getting started
-1. Use the `⚙️ Configure data` menu to select a datetime range or a dataset.
-2. Use the sidebar widgets to control how the data is split.
-3. Use the `📊 Folds` and `📈 Aggregations per fold` tabs to explore the effects.
-
-Please visit the [GitHub page](https://github.com/rsundqvist/time-split) for questions or feedback.
-
-#### ⚙️ Server configuration
-These values cannot be changed.
-
-{config.get_server_config_info()}
-
-See https://hub.docker.com/repository/docker/rsundqvist/time-split/ to spawn custom servers.
-
-### Server `time-split` version
-The server is using version `{__version__}`. To install this version, run
-
-```bash
-pip install time-split=={__version__}
-```
-
-in your favorite terminal. You may then copy one of the snippets from the `📊 Folds` tab to get started.
-"""  # noqa: S608  # This is not SQL.

time_split-0.5.0/src/time_split/streamlit/_views.py

@@ -1,125 +0,0 @@
-from pprint import pformat
-from typing import Any
-
-import pandas as pd
-import streamlit as st
-
-from time_split._compat import fmt_sec
-from time_split.streamlit import config
-from time_split.streamlit.widgets.display import CodeWidget, FoldOverviewWidget, PlotFoldsWidget
-from time_split.streamlit.widgets.types import QueryParams
-from time_split.support import create_explorer_link
-from time_split.types import DatetimeIndexSplitterKwargs, DatetimeSplits, DatetimeTypes
-
-
-def primary(
-    *,
-    df: pd.DataFrame,
-    plot_folds_widget: PlotFoldsWidget,
-    split_kwargs: DatetimeIndexSplitterKwargs,
-    limits: tuple[DatetimeTypes, DatetimeTypes],
-    dataset: str | None,
-    # Overview params
-    fold_overview_widget: FoldOverviewWidget,
-    splits: DatetimeSplits,
-    all_splits: DatetimeSplits,
-) -> None:
-    st.header("Folds", divider="rainbow")
-
-    with st.container(border=True):
-        left, right = st.columns([20, 4])
-        with right, st.container(border=True):
-            st.subheader("Plot preferences", divider=True)
-            plot_kwargs = plot_folds_widget.select()
-        with left:
-            plot_folds_widget.plot(split_kwargs, df, **plot_kwargs)
-
-    with st.container(border=True):
-        left, right = st.columns([20, 4])
-        with left:
-            st.subheader(
-                "Code snippets",
-                divider="rainbow",
-                help=(
-                    "The [integration](https://time-split.readthedocs.io/en/stable/generated/time_split.integration.html)"
-                    " modules accept the same parameters as `time_fold.split()`."
-                ),
-            )
-        with right, st.popover("🤝 Show permalink", use_container_width=True):
-            show_permalink(
-                split_kwargs=split_kwargs,
-                plot_kwargs=plot_kwargs,
-                limits=limits if dataset is None else dataset,
-            )
-
-        left, mid, right = st.columns([10, 10, 4])
-
-        with right:
-            with st.container(border=True):
-                st.subheader(
-                    "Type preferences",
-                    divider=True,
-                    help="The `time-split` package uses Pandas types internally. May not work for `📝 Free form` input.",
-                )
-                code_widget = CodeWidget.select()
-
-            st.write(
-                """
-                * Click [here](https://time-split.readthedocs.io/en/stable/generated/time_split.html#time_split.split) for `split()` docs.
-                * Click [here](https://time-split.readthedocs.io/en/stable/generated/time_split.html#time_split.plot) for `plot()` docs.
-                """
-            )
-
-            used, avail = fold_overview_widget.get_data_utilization(splits, limits)
-            st.caption(
-                f"Using `{fmt_sec(used)}` of `{fmt_sec(avail)}` (`{used / avail :.1%}`) of the available data range."
-            )
-
-        with left:
-            code_widget.show_split_code(split_kwargs, limits=limits)
-            fold_overview_widget.show_overview(splits, all_splits=all_splits)
-        with mid:
-            code_widget.show_plot_code(split_kwargs, plot_kwargs=plot_kwargs, limits=limits)
-
-
-def show_permalink(
-    *,
-    split_kwargs: DatetimeIndexSplitterKwargs,
-    plot_kwargs: dict[str, Any],
-    limits: tuple[DatetimeTypes, DatetimeTypes] | str,
-) -> None:
-    permalink_kwargs = {**split_kwargs, **plot_kwargs, "data": limits}
-    permalink_kwargs.pop("bar_labels")  # Not supported
-
-    host = config.PERMALINK_BASE_URL
-    if host == "":
-        host = "http://localhost:8501"
-        st.warning(f"May not be accurate; {config.PERMALINK_BASE_URL=} not set.", icon="⚠️")
-
-    link = create_explorer_link(host, **permalink_kwargs)
-    st.header("Share this link", divider=True)
-    with st.container(border=True):
-        st.write(f"[{link}]({link})")
-
-    convert = CodeWidget("string").convert
-
-    st.header(
-        "Parameters",
-        divider=True,
-        help="Input parameters are extracted from the URL in the address bar of your browser."
-        " Output parameters are used to generate the new link above.",
-    )
-    with st.container(border=True):
-        left, right = st.columns(2)
-        with left:
-            st.write("Input parameters.")
-            st.code(pformat(convert(QueryParams.get().to_dict(filter=False)), width=35))
-
-        with right:
-            st.write("Output parameters.")
-            st.code(pformat(convert(permalink_kwargs.copy()), width=35))
-        doc = "https://time-split.readthedocs.io/en/stable/generated/time_split.support.html#time_split.support.create_explorer_link"
-        st.caption(
-            f"Parameters may be [converted]({doc}) to equivalent values. "
-            "Note that `data` is called `available` by core library."
-        )

time_split-0.5.0/src/time_split/streamlit/config.py

@@ -1,68 +0,0 @@
-"""Configuration namespace.
-
-Values in this module are treated as hard limits by the server.
-"""
-
-from os import environ as _environ
-
-
-def _read_bool(key: str, default: bool) -> bool:
-    value = _environ.get(key, "").lower()
-    return value != "false" if default is True else value == "true"
-
-
-# Configurable by application users
-PLOT_RAW_TIMESERIES: bool = _read_bool("PLOT_AGGREGATIONS_PER_FOLD", True)
-"""Enable plot in the `🔍 Show raw data` tab."""
-PLOT_AGGREGATIONS_PER_FOLD: bool = _read_bool("PLOT_AGGREGATIONS_PER_FOLD", True)
-"""Enable plots in the `📈 Aggregations per fold` tab."""
-FIGURE_DPI: int = int(_environ.get("FIGURE_DPI", 200))
-"""Controls figure fidelity. Higher values look better, but is a lot slower."""
-MAX_SPLITS = int(_environ.get("MAX_SPLITS", 100))
-"""Upper fold count limit. Prevents figures from getting too large."""
-
-# Server configuration
-RAW_DATA_SAMPLES: int = int(_environ.get("RAW_DATA_SAMPLES", 1000))
-"""Maximum number of display and plot in `🔍 Show raw data` tab."""
-DATASETS_CONFIG_PATH: str = _environ.get("DATASETS_CONFIG_PATH") or "datasets.toml"
-"""Determines where to look for the dataset configuration TOML. Disable the dataset view if not found."""
-REQUIRE_DATASETS: bool = _read_bool("REQUIRE_DATASETS", False)
-"""If set, the server will refuse to start if not the ``DATASETS_CONFIG_PATH`` file does not exist."""
-DATASET_CACHE_TTL: int = int(_environ.get("DATASET_CACHE_TTL", 60 * 60))
-"""Cache timeout in seconds. Default is one hour."""
-PROCESS_QUERY_PARAMS: bool = _environ.get("PROCESS_QUERY_PARAMS", "").lower() != "false"
-"""Abort if parameters are given when ``False``. See :class:`~.widgets.types.QueryParams` for details."""
-PERMALINK_BASE_URL: str = _environ.get("PERMALINK_BASE_URL", "")
-"""Public base address for the application. Used to create permalinks."""
-
-
-def get_values() -> dict[str, int | bool]:
-    """Get config values as a dict."""
-    return {
-        "FIGURE_DPI": FIGURE_DPI,
-        "PLOT_RAW_TIMESERIES": PLOT_RAW_TIMESERIES,
-        "PLOT_AGGREGATIONS_PER_FOLD": PLOT_AGGREGATIONS_PER_FOLD,
-        "MAX_SPLITS": MAX_SPLITS,
-    }
-
-
-def get_descriptions() -> dict[str, str]:
-    """Get config key descriptions."""
-    return {
-        "FIGURE_DPI": "Controls figure fidelity. Higher values look better, but are slower to draw.",
-        "PLOT_RAW_TIMESERIES": "Enable plot in the `🔍 Show raw data` tab.",
-        "PLOT_AGGREGATIONS_PER_FOLD": "Enable plots in the `📈 Aggregations per fold` tab.",
-        "MAX_SPLITS": "Upper fold count limit. Prevents figures from getting too large.",
-    }
-
-
-def get_server_config_info() -> str:
-    """Get the readonly server configuration."""
-    description = get_descriptions()
-
-    return f"""
-* `{MAX_SPLITS=}`: {description['MAX_SPLITS']}
-* `{PLOT_RAW_TIMESERIES=}`: {description['PLOT_RAW_TIMESERIES']}
-* `{PLOT_AGGREGATIONS_PER_FOLD=}`: {description['PLOT_AGGREGATIONS_PER_FOLD']}
-* `{FIGURE_DPI=}`: {description['FIGURE_DPI']}
-"""

time_split-0.5.0/src/time_split/streamlit/widgets/__init__.py

@@ -1 +0,0 @@
-"""Streamlit components library."""

time_split-0.5.0/src/time_split/streamlit/widgets/data/__init__.py

@@ -1,7 +0,0 @@
-"""Widgets for interacting with datasets."""
-
-from ._data import DataWidget
-from ._datasets import DatasetWidget
-from ._sample_data import SampleDataWidget
-
-__all__ = ["DataWidget", "DatasetWidget", "SampleDataWidget"]