Skip to content

Commit

Permalink
Merge pull request #176 from 14Richa/14Richa/dashboard-elexonpy
Browse files Browse the repository at this point in the history 
Add elexonpy to requirements and display forecasts for different process types
  • Loading branch information
@peterdudfield
peterdudfield authored Aug 15, 2024
2 parents 88133ef + a20f718 commit df88c38
Show file tree
Hide file tree
Showing 4 changed files with 286 additions and 52 deletions.
2 changes: 1 addition & 1 deletion requirements.txt
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,4 +18,4 @@ fsspec==2024.2.0
s3fs==2024.2.0
ocf_blosc2==0.0.4
zarr==2.14.2

elexonpy
86 changes: 35 additions & 51 deletions src/forecast.py
View file
Original file line number Diff line number Diff line change
@@ -1,7 +1,7 @@
import os
from datetime import datetime, timedelta, time, timezone
from datetime import date, datetime, timedelta, time, timezone
import numpy as np

import pandas as pd
import plotly.graph_objects as go
import streamlit as st
from nowcasting_datamodel.connection import DatabaseConnection
Expand All @@ -14,12 +14,17 @@
from nowcasting_datamodel.read.read_gsp import get_gsp_yield, get_gsp_yield_sum

from plots.utils import (
get_colour_from_model_name, model_is_probabilistic, get_recent_available_model_names
get_colour_from_model_name,
model_is_probabilistic,
get_recent_available_model_names,
)

from plots.elexon_plots import add_elexon_plot


class GSPLabeler:
"""A function class to add the GSP name to the GSP IDs"""

def __init__(self, gsp_ids, gsp_names):
"""A function class to add the GSP name to the GSP IDs"""
self.gsp_ids = gsp_ids
Expand All @@ -29,78 +34,64 @@ def __call__(self, gsp_id):
"""Get GSP label"""
i = self.gsp_ids.index(gsp_id)
return f"{gsp_id}: {self.gsp_names[i]}"


def forecast_page():
"""Main page for status"""

st.markdown(
f'<h1 style="color:#63BCAF;font-size:48px;">{"National and GSP Forecasts"}</h1>',
unsafe_allow_html=True,
)

st.sidebar.subheader("Select Forecast Model")

connection = DatabaseConnection(url=os.environ["DB_URL"], echo=True)
with connection.get_session() as session:

# Add dropdown to select GSP region
locations = get_all_locations(session=session)
locations = [Location.from_orm(loc) for loc in locations if loc.gsp_id < 318]
gsp_ids = [loc.gsp_id for loc in locations]
gsp_names = [loc.region_name for loc in locations]

gsp_labeler = GSPLabeler(gsp_ids, gsp_names)

gsp_id = st.sidebar.selectbox(
"Select a region",
gsp_ids,
index=0,
format_func=gsp_labeler
)


gsp_id = st.sidebar.selectbox("Select a region", gsp_ids, index=0, format_func=gsp_labeler)

# Get effective capacity of selected GSP
capacity_mw = locations[gsp_ids.index(gsp_id)].installed_capacity_mw

# Find recent available models
available_models = get_recent_available_model_names(session)

# Add selection for models
selected_models = st.sidebar.multiselect("Select models", available_models, ["pvnet_v2"])

# If any selected models are probabilistic add checkbox to show quantiles
selected_prob_models = [model for model in selected_models if model_is_probabilistic(model)]
if len(selected_prob_models)>0:

if len(selected_prob_models) > 0:
show_prob = st.sidebar.checkbox("Show Probabilities Forecast", value=False)
else:
show_prob = False
if gsp_id!=0 and ("National_xg" in selected_models):

if gsp_id != 0 and ("National_xg" in selected_models):
selected_models.remove("National_xg")
st.sidebar.warning("National_xg only available for National forecast.")

# Add selection for adjuster
use_adjuster = st.sidebar.radio("Use adjuster", [True, False], index=1)

# Add selection for forecast type
forecast_type = st.sidebar.radio(
"Forecast Type", ["Now", "Creation Time", "Forecast Horizon"], index=0
)

now = datetime.now()
today = now.date()
yesterday = today - timedelta(days=1)

if forecast_type == "Now":
start_datetimes = [today - timedelta(days=2)]
end_datetimes = [None]

elif forecast_type == "Creation Time":
# Add calendar to select start date - defaults to yesterday

date_sel = st.sidebar.date_input("Forecast creation date:", yesterday)

# Add dropdown selection of init-times
dt_sel = datetime.combine(date_sel, time(0, 0))
initial_times = [dt_sel - timedelta(days=1) + timedelta(hours=3 * i) for i in range(8)]
Expand All @@ -111,7 +102,7 @@ def forecast_page():
initial_times,
[initial_times[x] for x in [14, 20, 26, 32, 38]],
)

select_init_times = sorted(select_init_times)

start_datetimes = select_init_times
Expand All @@ -121,7 +112,7 @@ def forecast_page():
# Add calendar and time selections for datetime
date_sel = st.sidebar.date_input("Forecast start date:", yesterday)
time_sel = st.sidebar.time_input("Forecast start time", time(0, 0))

dt_sel = datetime.combine(date_sel, time_sel)
start_datetimes = [dt_sel]
end_datetimes = [dt_sel + timedelta(days=2)]
Expand All @@ -133,21 +124,16 @@ def forecast_page():
list(range(0, 480, 30)) + list(range(480, 36 * 60, 180)),
0,
)

# Get the data to plot
forecast_per_model = {}
for model in selected_models:
for start_dt, end_dt in zip(start_datetimes, end_datetimes):

if forecast_type == "Now":
forecast_values = get_forecast_values_latest(
session=session,
gsp_id=gsp_id,
model_name=model,
start_datetime=start_dt,
session=session, gsp_id=gsp_id, model_name=model, start_datetime=start_dt
)
label = model

elif forecast_type == "Creation Time":
forecast_values = get_forecast_values(
session=session,
Expand All @@ -158,7 +144,7 @@ def forecast_page():
only_return_latest=True,
)
label = f"{model} {start_dt}"

elif forecast_type == "Forecast Horizon":
forecast_values = get_forecast_values(
session=session,
Expand All @@ -179,13 +165,10 @@ def forecast_page():
if use_adjuster:
forecast_value = forecast_value.adjust(limit=1000)
forecast_per_model[label].append(forecast_value)


# Get pvlive values
pvlive_data, pvlive_gsp_sum_dayafter, pvlive_gsp_sum_inday = get_pvlive_data(
end_datetimes[0], gsp_id, session, start_datetimes[0]
)

# Make figure
fig = go.Figure(
layout=go.Layout(
Expand All @@ -195,7 +178,6 @@ def forecast_page():
legend=go.layout.Legend(title=go.layout.legend.Title(text="Chart Legend")),
)
)

# Plot PVLive values and the forecasts
plot_pvlive(fig, gsp_id, pvlive_data, pvlive_gsp_sum_dayafter, pvlive_gsp_sum_inday)
plot_forecasts(fig, forecast_per_model, selected_prob_models, show_prob)
Expand All @@ -213,6 +195,9 @@ def forecast_page():
)
)

# Call the function to add Elexon plot and capture the returned figure
fig = add_elexon_plot(fig, start_datetimes, end_datetimes)

st.plotly_chart(fig, theme="streamlit")


Expand All @@ -227,7 +212,7 @@ def plot_pvlive(fig, gsp_id, pvlive_data, pvlive_gsp_sum_dayafter, pvlive_gsp_su
line["dash"] = "dash"

fig.add_trace(go.Scatter(x=x, y=y, mode="lines", name=k, line=line))

# pvlive gsp sum dictionary of values and chart for national forecast
if gsp_id == 0:
pvlive_gsp_sum_data = {}
Expand All @@ -241,7 +226,7 @@ def plot_pvlive(fig, gsp_id, pvlive_data, pvlive_gsp_sum_dayafter, pvlive_gsp_su
for k, v in pvlive_gsp_sum_data.items():
x = [i.datetime_utc for i in v]
y = [i.solar_generation_kw / 1000 for i in v]

line = {"color": get_colour_from_model_name(k)}
if k == "PVLive GSP Sum Estimate":
line["dash"] = "dash"
Expand Down Expand Up @@ -273,7 +258,7 @@ def plot_forecasts(fig, forecast_per_model, selected_prob_models, show_prob):
y=y,
mode="lines",
name=model,
line=dict(color=get_colour_from_model_name(model)),
line=dict(color=get_colour_from_model_name(model)),
opacity=opacity,
hovertemplate="<br>%{x}<br>" + "<b>%{y:.2f}</b>MW",
legendgroup=model,
Expand All @@ -300,7 +285,7 @@ def plot_forecasts(fig, forecast_per_model, selected_prob_models, show_prob):
showlegend=False,
)
)

fig.add_trace(
go.Scatter(
x=x,
Expand All @@ -313,8 +298,7 @@ def plot_forecasts(fig, forecast_per_model, selected_prob_models, show_prob):
showlegend=False,
)
)



except Exception as e:
print(e)
print("Could not add plevel to chart")
Expand Down
Loading

0 comments on commit df88c38

Please sign in to comment.