rainviewer_api.py

#!/usr/bin/env python
# coding: utf-8

# # Berlin Rain Map with RAINVIEWER

# In[ ]:





# In[142]:


import requests

from geopy.geocoders import Nominatim

from timezonefinder import TimezoneFinder
from datetime import datetime, timedelta
import pytz

from PIL import Image, ImageDraw, ImageFont, ImageSequence
from io import BytesIO

import glob
import os

import pandas as pd
# import geopandas as gpd  # isn't used
import plotly.express as px


# In[49]:


# import matplotlib.pyplot as plt
# import geoplot as gplt
# import geoplot.crs as gcrs
# import pyproj


# In[77]:





# ## 1. Getting the coordinates

# In[2]:


# input the name of the city

location_name = "Berlin"


# In[47]:


location = Nominatim(user_agent="RainRadar").geocode(location_name) # user_agent="geoapiExercises"
my_coordinates = [location.latitude, location.longitude]
# my_coordinates = [52.5170365, 13.3888599]

print("Location address:", location_name)
print("Latitude and Longitude of the said address:", (my_coordinates[0], my_coordinates[1]))


# In[5]:


obj = TimezoneFinder()
tz_value = obj.timezone_at(lat=my_coordinates[0], lng=my_coordinates[1])
tz_value = pytz.timezone(tz_value)

print(tz_value)


# ## 2. Conneting to API

# In[ ]:


# https://www.rainviewer.com/api/weather-maps-api.html


# In[135]:


weather_url = "https://api.rainviewer.com/public/weather-maps.json"
weather_response = requests.get(weather_url)
weather_response.status_code


# In[136]:


# weather_response.text
weather_request = weather_response.request
weather_request.headers


# In[137]:


# weather_response.json()


# In[138]:


print("Time of request:")
request_time = datetime.fromtimestamp(weather_response.json()['generated'], tz=tz_value)
print(request_time.strftime("%Y-%m-%d %H:%M:%S %Z"))


# ## 3. Plotting the geomap

# In[123]:


my_zoom = 4


# In[132]:


cities = pd.read_csv("worldcities.csv")
cities_coord = cities[['city', 'lat', 'lng', 'iso2', 'population']].sort_values('population', ascending=False)
cities_coord.head()

# cities_coord = gpd.GeoDataFrame(
#     cities_coord, geometry=gpd.points_from_xy(cities_coord.lng, cities_coord.lat), crs="EPSG:4326"
# )
# cities_coord.head()


# In[79]:


center_dict = {'lat': my_coordinates[0], 'lon': my_coordinates[1]}


# In[122]:


back_dir = "background_png"

# os.mkdir(back_dir)


# In[133]:


fig = px.scatter_mapbox(cities_coord[cities_coord['population'] >= 500000], 
                        lat="lat", lon="lng",
                        color_discrete_sequence=["darkblue"], # "fuchsia"
                        zoom=my_zoom, height=512, width=512,
                        center=center_dict,
                        size='population', size_max=20)
fig.update_layout(mapbox_style="open-street-map")
fig.update_layout(margin={"r":0,"t":0,"l":0,"b":0})
# fig.show()
fig.write_image(back_dir + "/" + str.lower(location_name) + "_zoom_" + str(my_zoom) + ".png")


# ## 4. Saving images and creating a gif

# In[83]:


# function: 

def download_png_by_url(weather_response, my_coordinates, size, zoom, color, smooth, snow, key, save_flg):
    '''
    Parameters:
        weather_response: response from https://api.rainviewer.com/public/weather-maps.json
        my_coordinates: list of 2 float elements (latitude, longitude)
        size: 256, 512
        zoom: 0, 1, 2, ...
        color: meanings see https://www.rainviewer.com/api/color-schemes.html
        smooth: 0 (don't blur radar data), 1 (blur radar data)
        snow: 0 (don't display snow), 1 (display snow in separate colors on the tiles)
        key: 'past', 'nowcast'
        save_flg: 0 (don't save the image in the folder), 1 (save)
    '''
    for dd in weather_response.json()['radar'][key]:
        img_url = (weather_response.json()['host'] + dd['path'] 
                   + '/' + str(size)
                   + '/' + str(zoom) 
                   + '/' + '/'.join(list(map(str, my_coordinates))) 
                   + '/' + str(color) + 
                   '/' + str(smooth) + '_' + str(snow) + 
                   '.png')
#         print(img_url)
        dttm = datetime.fromtimestamp(dd['time'], tz=tz_value) # pytz.utc
#         print(dttm.strftime("%Y-%m-%d %H:%M:%S"))
        img_response = requests.get(img_url)
        img = Image.open(BytesIO(img_response.content))
        
        draw = ImageDraw.Draw(img) 
        text1 = f"{dttm:%Y-%m-%d %H:%M:%S %Z}"
#         font1 = ImageFont.truetype('arial.ttf', 280) # 'Cantarell-VF.otf' 'Inconsolata-Light.ttf' 'Sacramento-Regular.ttf'
        position = (10, 10)
        left, top, right, bottom = draw.textbbox(position, text1) # , font=font1
        draw.rectangle((left-5, top-5, right+5, bottom+5), fill="lightgrey")
        draw.text(position, text=text1, fill="black") # , font=font1
#         img.show() # instead of display(img)
#         display(img) # doesn't work in GitHub Actions

        if save_flg == 1:
            if key == 'past':
                dir_name = png_past_dir
            elif key == 'nowcast':
                dir_name = png_nowcast_dir
            else:
                break
            img = img.save(dir_name + "/" + str.lower(location_name) + "_{}.png".format(dd['time']))
            
#         print("OK:", dttm.strftime("%Y-%m-%d %H:%M:%S %Z"))
        print("Done:", text1)
    return 1
    


# In[34]:


# function: deleting all files in given directory

def clear_folder(folder):
    files = glob.glob(folder + '/*')
    for f in files:
        os.remove(f)


# In[143]:


def filename_key(x):
    return int(x.split('_')[-1].split('.')[0])

# check if timestamp is not more than 2 days from request time
def filter_by_dttm(x):
#     datetime.timedelta(days=2, seconds=0, microseconds=0, milliseconds=0, minutes=0, hours=0, weeks=0)
    if request_time - datetime.fromtimestamp(filename_key(x), tz=tz_value) <= timedelta(days=2):
        return True
    else:
        return False
    
def make_gif(source_folder, target_folder):
    frames = [Image.open(image) for image in sorted(glob.glob(f"{source_folder}/*.png"), key=lambda x: filename_key(x)) if filter_by_dttm(image) is True]
    frame_one = frames[0]
    frame_one.save(target_folder + "/" + str.lower(location_name) + f"{request_time: %Y-%m-%d %H:%M:%S %Z}.gif", format="GIF", append_images=frames,
               save_all=True, duration=300, loop=0, transparency=0, disposal=2)

# disposal=2 - for replacement instead of overlaying 


# In[119]:


def gif_merge_back(gif_path, back_path, output_folder):
    gif_im = Image.open(gif_path)
    background = Image.open(back_path)
    frames = []
    for im_frame in ImageSequence.Iterator(gif_im):
        frame = background.copy()
        frame.paste(im_frame, mask=im_frame.convert("LA"))
        frames.append(frame)
    gif_name = gif_path.split('/')[-1].split('.')[0]
    frames[0].save(output_folder + "/" + gif_name + "_back.gif", save_all=True, duration=300, append_images=frames[1:], loop=0)
    


# In[62]:


# defining directories for saving images and gifs

png_past_dir = 'past_png'
png_nowcast_dir = 'nowcast_png'
gif_dir = 'past_gif'

# os.mkdir(png_past_dir)
# os.mkdir(png_nowcast_dir)
# os.mkdir(gif_dir)


# In[139]:


# saving images for past timestamps

download_png_by_url(weather_response, my_coordinates, 
                    size=512, zoom=my_zoom, color=4, smooth=0, snow=0, key='past', save_flg=1)


# In[140]:


# saving forecast images

download_png_by_url(weather_response, my_coordinates, 
                    size=512, zoom=my_zoom, color=4, smooth=0, snow=0, key='nowcast', save_flg=1)


# In[144]:


# creating a new gif file instead of a previous gif

clear_folder(gif_dir)
make_gif(png_past_dir, gif_dir)


# In[145]:


# adding the map background under the rain mask
                    
gif_path = [f for f in glob.glob(gif_dir + '/*.gif')][0]
back_path = [f for f in glob.glob(back_dir + '/*.png')][0]
output_folder = gif_dir

gif_merge_back(gif_path, back_path, output_folder)


# # ====== NOTES:

# In[ ]:


# Radar Object

# KEY			DESCRIPTION													VALUES
# past			Past weather radar frames. 2 hours, with 10-minute steps.	Array(Frame Object)	 
# nowcast		Future weather radar frames. 30 minutes.					Array(Frame Object)	 


# In[ ]:


# Frame Object

# KEY	DESCRIPTION	VALUES	EXAMPLE
# time	Map frame generation data in UNIX timestamp format (UTC). The map frame typically contains the images (radar, satellite) from different times, so this is not the time of the data rather than frame generation time.	Int(8)	1609401600
# path	Base path for the images of that frame. For information on its usage, refer to the next How to use host and path information section of this page	String(255)	/v2/satellite/0680143a9470


# In[ ]:


# {path}/{size}/{z}/{x}/{y}/{color}/{options}.png
# Radar data: displays one tile with the composite radar reflectivity data, with specified size, color scheme, and additional options.

# {path}/{size}/{z}/{lat}/{lon}/{color}/{options}.png
# Radar data: same as the link above, but with the center at specified coordinates (EPSG:4326) with desired zoom size. Great for widgets

# {path}/{big_size}/{color}/{options}.png
# Composite image with radar reflectivity for the entire world. Generates slowly, up to 10 seconds per image loads. Cannot be smoothed. IMPORTANT: color scheme and the snow mask for satellite data should always be “0”.



# In[ ]:


# {ts} – one of the available Unix timestamps from the API.
# {x}, {y}, {z} – x, y, and zoom level of the tile that you want to download. Read more about tiles
# {latitude}, {longitude} - latitude and longitude of specific coordinates accordingly. 
#     Decimal format. Must contain a dot in the number. Example: -32.7892, 108.67821.
# {size} – image size, can be 256 or 512.
# {big_size} – can be 2000, 4000, 8000, 16000, or 24000 (24000x12000 px or ~2km per pixel). 
#     For square images use 4096, 8196, and 16384 accordingly.
# {color} - the number of the color scheme from the provided list.
# {options} – list of options separated by the _ symbol. For example: 
#     ‘1_0’ means smoothed (1) image without snow color scheme (0). Now two options are available: {smooth}_{snow}
#     {smooth} - blur (1) or not (0) radar data. Large composite images are always not smoothed due to performance issues.
#     {snow} - display (1) or not (0) snow in separate colors on the tiles.


# In[ ]:


# Radar data: 
    
#     displays one tile with the composite radar reflectivity data, 
#     with specified size, color scheme, and additional options.
#     with the center at specified coordinates (EPSG:4326) with desired zoom size.