import os

import streamlit as st

import pandas as pd

import numpy as np

from numpy import ma

import plotly.graph_objects as go

from plotly.subplots import make_subplots

import plotly.io as pio

from datetime import datetime, timedelta

from dateutil import tz

import sqlite3

from sqlite3 import Connection

import plotly.express as px

from sklearn.preprocessing import normalize

from suntime import Sun

from utils.helpers import get_settings

profile = False

if profile:

try:

from pyinstrument import Profiler

except ImportError as e:

print(e)

profile = False

else:

profiler = Profiler()

profiler.start()

pio.templates.default = "plotly_white"

userDir = os.path.expanduser('~')

URI_SQLITE_DB = userDir + '/BirdNET-Pi/scripts/birds.db'

st.set_page_config(layout='wide')

# Remove whitespace from the top of the page

st.markdown("""

<style>

.css-18e3th9 {

padding-top: 2.5rem;

padding-bottom: 10rem;

padding-left: 5rem;

padding-right: 5rem;

}

.css-1d391kg {

padding-top: 3.5rem;

padding-right: 1rem;

padding-bottom: 3.5rem;

padding-left: 1rem;

}

</style>

""", unsafe_allow_html=True)

@st.cache_resource()

def get_connection(path: str):

uri = f"file:{path}?mode=ro"

return sqlite3.connect(uri, uri=True, check_same_thread=False)

def get_data(_conn: Connection):

df1 = pd.read_sql("SELECT * FROM detections", con=conn)

return df1

conn = get_connection(URI_SQLITE_DB)

df2 = get_data(conn)

df2['DateTime'] = pd.to_datetime(df2['Date'] + " " + df2['Time'])

df2 = df2.set_index('DateTime')

if len(df2) == 0:

st.info('No data yet. Please come back later.')

exit(0)

daily = st.sidebar.checkbox('Single Day View', help='Select if you want single day view, unselect for multi-day views')

if daily:

Start_Date = pd.to_datetime(df2.index.min()).date()

End_Date = pd.to_datetime(df2.index.max()).date()

end_date = st.sidebar.date_input('Date to View',

min_value=Start_Date,

max_value=End_Date,

value=(End_Date),

help='Select date for single day view')

start_date = end_date

else:

Start_Date = pd.to_datetime(df2.index.min()).date()

End_Date = pd.to_datetime(df2.index.max()).date()

start_date, end_date = st.sidebar.slider('Date Range',

min_value=Start_Date-timedelta(days=1),

max_value=End_Date,

value=(Start_Date, End_Date),

help='Select start and end date, if same date get a clockplot for a single day')

@st.cache_data()

def date_filter(df, start_date, end_date):

filt = (df2.index >= pd.Timestamp(start_date)) & (df2.index <= pd.Timestamp(end_date + timedelta(days=1)))

df = df[filt]

return (df)

df2 = date_filter(df2, start_date, end_date)

st.write('<style>div.row-widget.stRadio > div{flex-direction:row;justify-content: left;} </style>',

unsafe_allow_html=True)

st.write('<style>div.st-bf{flex-direction:column;} div.st-ag{font-weight:bold;padding-left:2px;}</style>',

unsafe_allow_html=True)

# Select time period buttons

# Disallow "Daily time period" for "Daily Chart"

if start_date == end_date:

resample_sel = st.sidebar.radio(

"Resample Resolution",

('Raw', '15 minutes', 'Hourly'), index=1, help='Select resolution for single day - larger times run faster')

resample_times = {'Raw': 'Raw',

'1 minute': '1min',

'15 minutes': '15min',

'Hourly': '1H'

}

resample_time = resample_times[resample_sel]

else:

resample_sel = st.sidebar.radio(

"Resample Resolution",

('Raw', '15 minutes', 'Hourly', 'DAILY'), index=1, help='Select resolution for species - DAILY provides time series')

resample_times = {'Raw': 'Raw',

'1 minute': '1min',

'15 minutes': '15min',

'Hourly': '1H',

'DAILY': '1D'

}

resample_time = resample_times[resample_sel]

@st.cache_data()

def time_resample(df, resample_time):

if resample_time == 'Raw':

df_resample = df['Com_Name']

else:

df_resample = df.resample(resample_time)['Com_Name'].aggregate('unique').explode()

return (df_resample)

top_bird = df2['Com_Name'].mode()[0]

df5 = time_resample(df2, resample_time)

# Create species count for selected date range

Specie_Count = df5.value_counts()

# Create Hourly Crosstab

hourly = pd.crosstab(df5, df5.index.hour, dropna=True, margins=True)

# Filter on species

species = list(hourly.sort_values("All", ascending=False).index)

if len(Specie_Count) > 1:

top_N = st.sidebar.slider(

'Select Number of Birds to Show',

min_value=1,

max_value=len(Specie_Count),

value=min(10, len(Specie_Count))

)

else:

top_N = 1

top_N_species = (df5.value_counts()[:top_N])

font_size = 15

def sunrise_sunset_scatter(date_range):

conf = get_settings()

latitude = conf.getfloat('LATITUDE')

longitude = conf.getfloat('LONGITUDE')

sun = Sun(latitude, longitude)

sunrise_list = []

sunset_list = []

sunrise_text_list = []

sunset_text_list = []

daysback_range = []

local_timezone = tz.tzlocal()

for current_date in date_range:

current_datetime = datetime.combine(current_date, datetime.min.time())

sun_rise = sun.get_sunrise_time(current_datetime, local_timezone)

sun_dusk = sun.get_sunset_time(current_datetime, local_timezone)

sun_rise_time = float(sun_rise.hour) + float(sun_rise.minute) / 60.0

sun_dusk_time = float(sun_dusk.hour) + float(sun_dusk.minute) / 60.0

temp_time = str(sun_rise)[-14:-9] + " Sunrise"

sunrise_text_list.append(temp_time)

temp_time = str(sun_dusk)[-14:-9] + " Sunset"

sunset_text_list.append(temp_time)

sunrise_list.append(sun_rise_time)

sunset_list.append(sun_dusk_time)

daysback_range.append(current_date.strftime('%d-%m-%Y'))

sunrise_list.append(None)

sunrise_text_list.append(None)

sunrise_list.extend(sunset_list)

sunrise_text_list.extend(sunset_text_list)

daysback_range.append(None)

daysback_range.extend(daysback_range)

return daysback_range, sunrise_list, sunrise_text_list

def hms_to_dec(t):

h = t.hour

m = t.minute / 60

s = t.second / 3600

result = h + m + s

return result

def hms_to_str(t):

h = t.hour

m = t.minute

return "%02d:%02d" % (h, m)

if daily is False:

if resample_time != '1D':

specie = st.selectbox(

'Which bird would you like to explore for the dates '

+ str(start_date) + ' to ' + str(end_date) + '?',

species,

index=0)

if specie == 'All':

df_counts = int(hourly[hourly.index == specie]['All'].iloc[0])

fig = make_subplots(

rows=3, cols=2,

specs=[[{"type": "xy", "rowspan": 3}, {"type": "polar", "rowspan": 2}],

[{"rowspan": 1}, {"rowspan": 1}],

[None, {"type": "xy", "rowspan": 1}]],

subplot_titles=('<b>Top ' + str(top_N) + ' Species in Date Range ' + str(start_date) + ' to ' + str(

end_date) + '<br>for ' + str(resample_sel) + ' sampling interval.' + '</b>',

'Total Detect:' + str('{:,}'.format(df_counts))

)

)

fig.layout.annotations[1].update(x=0.7, y=0.25, font_size=15)

# Plot seen species for selected date range and number of species

fig.add_trace(go.Bar(y=top_N_species.index.tolist(), x=top_N_species.values.tolist(), orientation='h', marker_color='seagreen'), row=1, col=1)

fig.update_layout(

margin=dict(l=0, r=0, t=50, b=0),

yaxis={'categoryorder': 'total ascending'})

# Set 360 degrees, 24 hours for polar plot

theta = np.linspace(0.0, 360, 24, endpoint=False).tolist()

specie_filt = df5 == specie

df3 = df5[specie_filt]

detections2 = pd.crosstab(df3, df3.index.hour)

d = pd.DataFrame(np.zeros((24, 1))).squeeze()

detections = hourly.loc[specie]

detections = (d + detections).fillna(0)

fig.add_trace(go.Barpolar(r=detections.tolist(), theta=theta, marker_color='seagreen'), row=1, col=2)

fig.update_layout(

autosize=False,

width=1000,

height=500,

showlegend=False,

polar=dict(

radialaxis=dict(

tickfont_size=font_size,

showticklabels=False,

hoverformat="#%{theta}: <br>Popularity: %{percent} </br> %{r}"

),

angularaxis=dict(

tickfont_size=font_size,

rotation=-90,

direction='clockwise',

tickmode='array',

tickvals=[0, 15, 35, 45, 60, 75, 90, 105, 120, 135, 150, 165, 180, 195, 210,

225, 240, 255, 270, 285, 300, 315, 330, 345],

ticktext=['12am', '1am', '2am', '3am', '4am', '5am', '6am', '7am', '8am', '9am',

'10am', '11am', '12pm', '1pm', '2pm', '3pm', '4pm', '5pm', '6pm',

'7pm', '8pm', '9pm', '10pm', '11pm'],

hoverformat="#%{theta}: <br>Popularity: %{percent} </br> %{r}"

),

),

)

daily = pd.crosstab(df5, df5.index.date, dropna=True, margins=True)

fig.add_trace(go.Bar(x=daily.columns[:-1].tolist(), y=daily.loc[specie][:-1].tolist(), marker_color='seagreen'), row=3, col=2)

st.plotly_chart(fig, use_container_width=True) # , config=config)

else:

col1, col2 = st.columns(2)

with col1:

fig = make_subplots(

rows=3, cols=1,

specs=[[{"type": "polar", "rowspan": 2}], [{"rowspan": 1}], [{"type": "xy", "rowspan": 1}]]

)

# Set 360 degrees, 24 hours for polar plot

theta = np.linspace(0.0, 360, 24, endpoint=False).tolist()

specie_filt = df5 == specie

df3 = df5[specie_filt]

detections2 = pd.crosstab(df3, df3.index.hour)

d = pd.DataFrame(np.zeros((24, 1))).squeeze()

detections = hourly.loc[specie]

detections = (d + detections).fillna(0)

fig.add_trace(go.Barpolar(r=detections.tolist(), theta=theta, marker_color='seagreen'), row=1, col=1)

fig.update_layout(

autosize=False,

width=1000,

height=500,

showlegend=False,

polar=dict(

radialaxis=dict(

tickfont_size=font_size,

showticklabels=False,

hoverformat="#%{theta}: <br>Popularity: %{percent} </br> %{r}"

),

angularaxis=dict(

tickfont_size=font_size,

rotation=-90,

direction='clockwise',

tickmode='array',

tickvals=[0, 15, 35, 45, 60, 75, 90, 105, 120, 135, 150, 165, 180, 195,

210, 225, 240, 255, 270, 285, 300, 315, 330, 345],

ticktext=['12am', '1am', '2am', '3am', '4am', '5am', '6am', '7am', '8am',

'9am', '10am', '11am', '12pm', '1pm', '2pm', '3pm', '4pm', '5pm',

'6pm', '7pm', '8pm', '9pm', '10pm', '11pm'],

hoverformat="#%{theta}: <br>Popularity: %{percent} </br> %{r}"

),

),

)

daily = pd.crosstab(df5, df5.index.date, dropna=True, margins=True)

fig.add_trace(go.Bar(x=daily.columns[:-1].tolist(), y=daily.loc[specie][:-1].tolist(), marker_color='seagreen'), row=3, col=1)

st.plotly_chart(fig, use_container_width=True) # , config=config)

df_counts = int(hourly[hourly.index == specie]['All'].iloc[0])

st.subheader('Total Detect:' + str('{:,}'.format(df_counts))

+ ' Confidence Max:' +

str('{:.2f}%'.format(max(df2[df2['Com_Name'] == specie]['Confidence']) * 100))

+ ' ' + ' Median:' +

str('{:.2f}%'.format(np.median(df2[df2['Com_Name'] == specie]['Confidence']) * 100)))

recordings = df2[df2['Com_Name'] == specie]['File_Name']

with col2:

try:

recording = st.selectbox('Available recordings', recordings.sort_index(ascending=False))

date_specie = df2.loc[df2['File_Name'] == recording, ['Date', 'Com_Name']]

date_dir = date_specie['Date'].values[0]

specie_dir = date_specie['Com_Name'].values[0].replace(" ", "_").replace("'", "")

st.image(userDir + '/BirdSongs/Extracted/By_Date/' + date_dir + '/' + specie_dir + '/' + recording + '.png')

st.audio(userDir + '/BirdSongs/Extracted/By_Date/' + date_dir + '/' + specie_dir + '/' + recording)

except Exception:

st.title('RECORDING NOT AVAILABLE :(')

else:

specie = st.selectbox('Which bird would you like to explore for the dates '

+ str(start_date) + ' to ' + str(end_date) + '?',

species[1:],

index=0)

df_counts = int(hourly.loc[hourly.index == specie, 'All'].iloc[0])

fig = make_subplots(rows=1, cols=1)

df4 = df2['Com_Name'][df2['Com_Name'] == specie].resample('15min').count()

df4.index = [df4.index.date, df4.index.time]

day_hour_freq = df4.unstack().fillna(0)

saved_time_labels = [hms_to_str(h) for h in day_hour_freq.columns.tolist()]

fig_dec_y = [hms_to_dec(h) for h in day_hour_freq.columns.tolist()]

fig_x = [d.strftime('%d-%m-%Y') for d in day_hour_freq.index.tolist()]

fig_y = [h.strftime('%H:%M') for h in day_hour_freq.columns.tolist()]

day_hour_freq.columns = fig_dec_y

fig_z = day_hour_freq.values.transpose().tolist()

color_pals = px.colors.named_colorscales()

selected_pal = st.sidebar.selectbox('Select Color Pallet for Daily Detections', color_pals)

heatmap = go.Heatmap(

x=fig_x,

y=day_hour_freq.columns.tolist(),

z=fig_z, # heat.values,

showscale=False,

texttemplate="%{text}", autocolorscale=False, colorscale=selected_pal

)

daysback_range, sunrise_list, sunrise_text_list = sunrise_sunset_scatter(day_hour_freq.index.tolist())

sunrise_sunset = go.Scatter(x=daysback_range,

y=sunrise_list,

mode='lines',

hoverinfo='text',

text=sunrise_text_list,

line_color='orange', line_width=1, name=' ')

fig = go.Figure(data=[heatmap, sunrise_sunset])

number_of_y_ticks = 12

y_downscale_factor = int(len(saved_time_labels) / number_of_y_ticks)

fig.update_layout(

yaxis=dict(

tickmode='array',

tickvals=day_hour_freq.columns[::y_downscale_factor],

ticktext=saved_time_labels[::y_downscale_factor],

nticks=6

)

)

st.plotly_chart(fig, use_container_width=True) # , config=config)

else:

fig = make_subplots(

rows=1, cols=2,

specs=[[{"type": "xy", "rowspan": 1}, {"type": "xy", "rowspan": 1}]],

subplot_titles=('<b>Top ' + str(top_N) + ' Species For ' + str(start_date) + '</b>',

'<b>Daily ' + str(start_date) + ' Detections on ' + resample_sel + ' interval</b>'),

shared_yaxes='all',

horizontal_spacing=0

)

df6 = df5.to_frame(name='Com_Name')

readings = top_N

plt_topN_today = (df6['Com_Name'].value_counts()[:readings])

freq_order = pd.value_counts(df6['Com_Name']).iloc[:readings].index

fig.add_trace(go.Bar(y=plt_topN_today.index.tolist(), x=plt_topN_today.values.tolist(), marker_color='seagreen', orientation='h'), row=1,

col=1)

df6['Hour of Day'] = [r.hour for r in df6.index.time]

heat = pd.crosstab(df6['Com_Name'], df6['Hour of Day'])

# Order heatmap Birds by frequency of occurrance

heat.index = pd.CategoricalIndex(heat.index, categories=freq_order)

heat.sort_index(level=0, inplace=True)

heat.index = heat.index.astype(str)

heat_plot_values = ma.log(heat.values).filled(0)

hours_in_day = pd.Series(data=range(0, 24))

heat_frame = pd.DataFrame(data=0, index=heat.index, columns=hours_in_day)

heat = (heat + heat_frame).fillna(0)

heat_values_normalized = normalize(heat.values, axis=1, norm='l1')

labels = heat.values.astype(int).astype('str')

labels[labels == '0'] = ""

fig.add_trace(go.Heatmap(x=heat.columns.tolist(), y=heat.index.tolist(), z=heat_values_normalized, # heat.values,

showscale=False,

text=labels, texttemplate="%{text}", colorscale='Blugrn'

), row=1, col=2)

fig.update_yaxes(visible=True, autorange="reversed", ticks="inside", tickson="boundaries", ticklen=10000,

showgrid=True)

fig.update_layout(xaxis_ticks="inside",

margin=dict(l=0, r=0, t=50, b=0))

st.plotly_chart(fig, use_container_width=True) # , config=config)

if profile:

profiler.stop()

profiler.print()

print('**profiler done**', flush=True)