JobTask¶

def download_quicklooks(
    self,
    output_directory: Union[str, Path, None] = None,
) -> List[str]:
    """
    Downloads quicklooks of the job task to disk.

    After download, can be plotted via jobtask.plot_quicklooks().

    Args:
        output_directory: The file output directory, defaults to the current working
            directory.

    Returns:
        The quicklooks filepaths.
    """
    if output_directory is None:
        # On purpose downloading the quicklooks to the jobs folder and not the
        # jobtasks folder,since only relevant for data block task. And clearer
        # for job.download_quicklooks.
        output_directory = (
            Path.cwd() / f"project_{self.auth.project_id}" / f"job_{self.job_id}"
        )
    else:
        output_directory = Path(output_directory)
    output_directory.mkdir(parents=True, exist_ok=True)
    logger.info(f"Download directory: {str(output_directory)}")

    url = (
        f"{self.auth._endpoint()}/projects/{self.project_id}/jobs/{self.job_id}"
        f"/tasks/{self.jobtask_id}/outputs/quicklooks/"
    )
    response_json = self.auth._request(request_type="GET", url=url)
    quicklooks_ids = response_json["data"]

    out_paths: List[str] = []
    for ql_id in tqdm(quicklooks_ids):
        out_path = output_directory / f"quicklook_{ql_id}"  # No suffix required.
        out_paths.append(str(out_path))

        url = (
            f"{self.auth._endpoint()}/projects/{self.project_id}/jobs/{self.job_id}"
            f"/tasks/{self.jobtask_id}/outputs/quicklooks/{ql_id}"
        )
        response = self.auth._request(
            request_type="GET", url=url, return_text=False
        )

        with open(out_path, "wb") as dst:
            for chunk in response:
                dst.write(chunk)

    self.quicklooks = out_paths  # pylint: disable=attribute-defined-outside-init
    return out_paths

def download_results(
    self, output_directory: Union[str, Path, None] = None
) -> List[str]:
    """
    Downloads and unpacks the jobtask results. Default download to Desktop.

    Args:
        output_directory: The file output directory, defaults to the current working
            directory.
    Returns:
        List of the downloaded results' filepaths.
    """
    logger.info(f"Downloading results of jobtask {self.jobtask_id}")

    if output_directory is None:
        output_directory = (
            Path.cwd()
            / f"project_{self.auth.project_id}/job_{self.job_id}/jobtask_{self.jobtask_id}"
        )
    else:
        output_directory = Path(output_directory)
    output_directory.mkdir(parents=True, exist_ok=True)
    logger.info(f"Download directory: {str(output_directory)}")

    download_url = self._get_download_url()
    out_filepaths = download_results_from_gcs(
        download_url=download_url,
        output_directory=output_directory,
    )

    self.results = out_filepaths
    return out_filepaths

def get_results_json(self, as_dataframe: bool = False) -> Union[Dict, GeoDataFrame]:
    """
    Gets the Jobtask results data.json.

    Args:
        as_dataframe: "fc" for FeatureCollection dict, "df" for GeoDataFrame.

    Returns:
        Json of the results, alternatively geodataframe.
    """
    url = (
        f"{self.auth._endpoint()}/projects/{self.auth.project_id}/jobs/{self.job_id}"
        f"/tasks/{self.jobtask_id}/outputs/data-json/"
    )
    response_json = self.auth._request(request_type="GET", url=url)
    logger.info(f"Retrieved {len(response_json['features'])} features.")

    if as_dataframe:
        # UP42 results are always in EPSG 4326
        df = GeoDataFrame.from_features(response_json, crs=4326)
        return df
    else:
        return response_json

def map_results(
    self,
    aoi: GeoDataFrame = None,
    show_images: bool = True,
    show_features: bool = True,
    name_column: str = "uid",
    save_html=None,
) -> folium.Map:
    """
    Displays data.json, and if available, one or multiple results geotiffs.

    Args:
        aoi: GeoDataFrame of aoi.
        show_images: Shows images if True (default).
        show_features: Shows features if True (default).
        name_column: Name of the feature property that provides the Feature/Layer name.
        save_html: The path for saving folium map as html file. With default None, no file is saved.
    """
    # TODO: Surface optional filepaths? or remove option alltogether?
    if self.results is None:
        raise ValueError(
            "You first need to download the results via job.download_results()!"
        )

    # Add features to map.
    # Some blocks store vector results in an additional geojson file.
    # pylint: disable=not-an-iterable
    json_fp = [fp for fp in self.results if fp.endswith(".geojson")]
    if json_fp:
        json_fp = json_fp[0]
    else:
        # pylint: disable=not-an-iterable
        json_fp = [fp for fp in self.results if fp.endswith(".json")][0]
    df: GeoDataFrame = gpd.read_file(json_fp)

    # Add image to map.
    m = self._map_images(
        plot_file_format=[".tif"],
        result_df=df,
        filepaths=self.results,
        aoi=aoi,
        show_images=show_images,
        show_features=show_features,
        name_column=name_column,
        save_html=save_html,
    )
    return m

def plot_quicklooks(
    self,
    figsize: Tuple[int, int] = (8, 8),
    filepaths: List = None,
    titles: List[str] = None,
) -> None:
    """
    Plots the downloaded quicklooks (filepaths saved to self.quicklooks of the
    respective object, e.g. job, catalog).

    Args:
            figsize: matplotlib figure size.
            filepaths: Paths to images to plot. Optional, by default picks up the last
                    downloaded results.
            titles: List of titles for the subplots, optional.

    """
    if filepaths is None:
        if self.quicklooks is None:
            raise ValueError("You first need to download the quicklooks!")
        filepaths = self.quicklooks

    warnings.filterwarnings(
        "ignore", category=rasterio.errors.NotGeoreferencedWarning
    )
    self.plot_results(
        plot_file_format=[".jpg", ".jpeg", ".png"],
        figsize=figsize,
        filepaths=filepaths,
        titles=titles,
    )

def plot_results(
    self,
    figsize: Tuple[int, int] = (14, 8),
    filepaths: Union[List[Union[str, Path]], Dict] = None,
    titles: List[str] = None,
    # pylint: disable=dangerous-default-value
    plot_file_format: List[str] = [".tif"],
) -> None:
    """
    Plots image data (quicklooks or results)

    Args:
        plot_file_format: List of accepted image file formats e.g. [".tif"]
        figsize: matplotlib figure size.
        filepaths: Paths to images to plot. Optional, by default picks up the last
            downloaded results.
        titles: Optional list of titles for the subplots.

    """
    if filepaths is None:
        if self.results is None:
            raise ValueError("You first need to download the results!")
        filepaths = self.results
        # Unpack results path dict in case of jobcollection.
        if isinstance(filepaths, dict):
            filepaths_lists = list(filepaths.values())
            filepaths = [item for sublist in filepaths_lists for item in sublist]

    if not isinstance(filepaths, list):
        filepaths = [filepaths]  # type: ignore
    filepaths = [Path(path) for path in filepaths]

    imagepaths = [
        path for path in filepaths if str(path.suffix) in plot_file_format  # type: ignore
    ]
    if not imagepaths:
        raise ValueError(
            f"This function only plots files of format {plot_file_format}."
        )

    if not titles:
        titles = [Path(fp).stem for fp in imagepaths]
    if not isinstance(titles, list):
        titles = [titles]  # type: ignore

    if len(imagepaths) < 2:
        nrows, ncols = 1, 1
    else:
        ncols = 3
        nrows = int(math.ceil(len(imagepaths) / float(ncols)))

    _, axs = plt.subplots(nrows=nrows, ncols=ncols, figsize=figsize)
    if len(imagepaths) > 1:
        axs = axs.ravel()
    else:
        axs = [axs]

    for idx, (fp, title) in enumerate(zip(imagepaths, titles)):
        with rasterio.open(fp) as src:
            img_array = src.read()[:3, :, :]
            # TODO: Handle more band configurations.
            # TODO: add histogram equalization?
            show(
                img_array,
                transform=src.transform,
                title=title,
                ax=axs[idx],
                aspect="auto",
            )
        axs[idx].set_axis_off()
    plt.axis("off")
    plt.tight_layout()
    plt.show()