UK NRFA (add more data)

[simonmoulds]

This relates to #5 and #32

The current UKFetcher access data from the EA Hydrology API, but in fact this API only provides access to England data. Here is a MWE to download data from the National River Flow Archive (NRFA), which is the official UK record of river flow, providing access to daily, monthly and peak flow data for around 1600 catchments across the UK.

import requests
import json
import pandas as pd

# The base URL to access the nrfa API
BASE_URL = "https://nrfaapps.ceh.ac.uk/nrfa/ws"

VALID_DATA_TYPES = [
    'gdf', 'ndf', 'gmf', 'nmf', 'cdr', 'cdr-d', 'cmr',
    'pot-stage', 'pot-flow', 'gauging-stage', 'gauging-flow',
    'amax-stage', 'amax-flow'
]


def catalogue():
    """Get the NRFA data catalog and metadata."""
    query_params = {
        "station": "*",
        "format": "json-object",
        "fields": "all"
    }
    # Send request using requests
    response = requests.get(f"{BASE_URL}/station-info", params=query_params)
    response.raise_for_status()  # raises an error for non-200 responses
    # Decode JSON and load into DataFrame
    data = response.json()
    df = pd.DataFrame(data["data"])
    return df


def _build_ts(response):
    """This is used to parse timeseries data from NRFA."""
    variable = response['data-type']['id']
    dates = response['data-stream'][0::2]
    values = response['data-stream'][1::2]
    df = pd.DataFrame.from_dict({'time': dates, variable: values})
    # Format `time` column as datetime. The API docs specify that
    # times returned by the API have ISO8601 format.
    df['time'] = pd.to_datetime(df['time'], format='ISO8601')
    return df


def get_ts(id, data_type):
    """Get a timeseries from the UK National River Flow Archive."""
    query_params = {
        "station": str(id),
        "data-type": data_type,
        "format": "json-object"
    }
    response = requests.get(f"{BASE_URL}/time-series", params=query_params)
    response.raise_for_status()
    data = response.json()
    df = _build_ts(data)
    return df

metadata = catalogue()
ts = get_ts(1001, "gdf")```

[kratzert]

Closed by #38