Skip to content

Latest commit

 

History

History
578 lines (493 loc) · 13.3 KB

README.md

File metadata and controls

578 lines (493 loc) · 13.3 KB
Raw

This is a python script to visualize the weather of 500+ cities across the world of varying distance from the equator. The objective is to build a series of scatter plots to showcase the following relationships:

  • Temperature (F) vs. Latitude
  • Humidity (%) vs. Latitude
  • Cloudiness (%) vs. Latitude
  • Wind Speed (mph) vs. Latitude

A minimum of 500 cities worldwide will be selected for this exercise based on latitude and longitude for weather visualization. for random selection of latitude and longitude, random module will be utilized, and then using citipy, cities nearest to the selected random latitude/longitude will be obtained.

import pandas as pd
from citipy import citipy
import matplotlib.pyplot as plt
%notebook inline
import random
from config import api_key
import openweathermapy.core as owm
#seedify random numbers to obtain repeatable results
random.seed(a=20)
city_list=[];country_list=[]
for i in range(0,1200):
    lat=random.randrange(-90,90)
    long=random.randrange(-180,180)
    city=citipy.nearest_city(lat,long)
    city_list.append(city.city_name)
    country_list.append(city.country_code)
df=pd.DataFrame({"City":city_list,"Country":country_list}) 
df.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
City Country
0 yellowknife ca
1 noumea nc
2 tuktoyaktuk ca
3 taman id
4 avarua ck

After obtaining 500 cities, it is important to check that there are no repeat cities. So, we would need to count all unique combinations of city and country. This is the result I got after various iterations, by manually increasing number of selection to obtain at least 500 unique cities.

  1. 268 unique combinations of city, country out of 500 random pickings
  2. 448 unique combinations of city, country out of 1000 random pickings
  3. 510 unique combinations of city, country out of 1200 random pickings Seed Number was not changed in these iterations.
# Counting unique combinations of city and country
grped=df.groupby(["Country","City"])
ds=grped.size()
len(ds)
510

df_=ds.to_frame().reset_index()
df_unique_cities=df_[["Country","City"]]
# Creating column with city,country in the format required by openweathermapy
df_unique_cities['city,country']=""
for index,row in df_unique_cities.iterrows():
    row['city,country']=f"{row['City']},{row['Country']}"
df_unique_cities['Lat']="";df_unique_cities['Long']="";df_unique_cities['Temperature(F)']="";df_unique_cities['Humidity(%)']=""
df_unique_cities['Cloudiness(%)']="";df_unique_cities['Wind Speed(mph)']=""
df_unique_cities.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
Country City city,country Lat Long Temperature(F) Humidity(%) Cloudiness(%) Wind Speed(mph)
0 ae sharjah sharjah,ae
1 af baraki barak baraki barak,af
2 ag codrington codrington,ag
3 ai the valley the valley,ai
4 ao lucapa lucapa,ao

Pulling data from openweathermapI pulled the latitude and longitude of cities to get the exact values

Pulling Data From openweathermap api

import urllib
settings = {"APPID": api_key, "units": "imperial"}
# keys = ["coord.lon","coord.lat","clouds.all","main.temp", "main.humidity", "wind.speed"]
for index,row in df_unique_cities.iterrows():
    try:
        data=owm.get_current(row['city,country'],**settings)
        row['Lat']=data("coord.lat");row['Long']=data("coord.lat")
        row['Temperature(F)']=data("main.temp");row['Humidity(%)']=data("main.humidity")
        row['Cloudiness(%)']=data("clouds.all");row['Wind Speed(mph)']=data("wind.speed")
        #     data.get_many(keys)
    except urllib.error.HTTPError as err:
        if err==404:
            print(row['city,country'])
            continue
        else:
            print(err)
            print(row['city,country'])
HTTP Error 404: Not Found
codrington,ag
HTTP Error 404: Not Found
ngukurr,au
HTTP Error 404: Not Found
acarau,br
HTTP Error 404: Not Found
igarape-miri,br
HTTP Error 404: Not Found
laguna,br
HTTP Error 404: Not Found
bokspits,bw
HTTP Error 404: Not Found
longlac,ca
HTTP Error 404: Not Found
louisbourg,ca
HTTP Error 404: Not Found
saint anthony,ca
HTTP Error 404: Not Found
lasa,cn
HTTP Error 404: Not Found
warqla,dz
HTTP Error 404: Not Found
kemijarvi,fi
HTTP Error 404: Not Found
illoqqortoormiut,gl
HTTP Error 404: Not Found
marathopolis,gr
HTTP Error 404: Not Found
bolungarvik,is
HTTP Error 404: Not Found
skagastrond,is
HTTP Error 404: Not Found
yomitan,jp
HTTP Error 404: Not Found
chardara,kz
HTTP Error 404: Not Found
karaton,kz
HTTP Error 404: Not Found
asfi,ma
HTTP Error 404: Not Found
mananara,mg
HTTP Error 404: Not Found
taolanaro,mg
HTTP Error 404: Not Found
tsihombe,mg
HTTP Error 404: Not Found
grand river south east,mu
HTTP Error 404: Not Found
canitas,mx
HTTP Error 404: Not Found
san quintin,mx
HTTP Error 404: Not Found
camana,pe
HTTP Error 404: Not Found
marcona,pe
HTTP Error 404: Not Found
mataura,pf
HTTP Error 404: Not Found
airai,pw
HTTP Error 404: Not Found
amderma,ru
HTTP Error 404: Not Found
belushya guba,ru
HTTP Error 404: Not Found
chagda,ru
HTTP Error 404: Not Found
ishlei,ru
HTTP Error 404: Not Found
kadykchan,ru
HTTP Error 404: Not Found
kamenskoye,ru
HTTP Error 404: Not Found
karaul,ru
HTTP Error 404: Not Found
khonuu,ru
HTTP Error 404: Not Found
krasnoselkup,ru
HTTP Error 404: Not Found
ksenyevka,ru
HTTP Error 404: Not Found
mys shmidta,ru
HTTP Error 404: Not Found
nizhneyansk,ru
HTTP Error 404: Not Found
primore,ru
HTTP Error 404: Not Found
sentyabrskiy,ru
HTTP Error 404: Not Found
shchelyayur,ru
HTTP Error 404: Not Found
tumannyy,ru
HTTP Error 404: Not Found
ust-kamchatsk,ru
HTTP Error 404: Not Found
sakakah,sa
HTTP Error 404: Not Found
barentsburg,sj
HTTP Error 404: Not Found
bargal,so
HTTP Error 404: Not Found
faya,td
HTTP Error 404: Not Found
asau,tv
HTTP Error 404: Not Found
karauzyak,uz
HTTP Error 404: Not Found
da nang,vn
HTTP Error 404: Not Found
halalo,wf
HTTP Error 404: Not Found
vaitupu,wf
HTTP Error 404: Not Found
saleaula,ws
HTTP Error 404: Not Found
samusu,ws
HTTP Error 404: Not Found
satitoa,ws
HTTP Error 404: Not Found
scottsburgh,za
HTTP Error 404: Not Found
umzimvubu,za

#changing the datatype from str to numeric
df_unique_cities[["Lat","Long","Temperature(F)","Humidity(%)","Cloudiness(%)","Wind Speed(mph)"]]=df_unique_cities[["Lat","Long","Temperature(F)","Humidity(%)","Cloudiness(%)",
                                                                                                                    "Wind Speed(mph)"]].apply(pd.to_numeric,errors="coerce")
df_unique_cities.count()
Country            510
City               510
city,country       510
Lat                449
Long               449
Temperature(F)     449
Humidity(%)        449
Cloudiness(%)      449
Wind Speed(mph)    449
dtype: int64

Hence, weather data of 449 (out of 510) was returned from openweathermap. We will utilize data of 449 cities for our final plots.

df_unique_cities=df_unique_cities.dropna(axis=0,how="any")
df_unique_cities.to_csv("Output/city_weatherdata.csv")
df_unique_cities.head()
<style scoped> .dataframe tbody tr th:only-of-type { vertical-align: middle; } 
.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}
</style>
Country City city,country Lat Long Temperature(F) Humidity(%) Cloudiness(%) Wind Speed(mph)
0 ae sharjah sharjah,ae 25.36 25.36 88.20 46.0 0.0 3.36
1 af baraki barak baraki barak,af 33.97 33.97 62.60 39.0 0.0 4.70
3 ai the valley the valley,ai 18.22 18.22 82.47 74.0 40.0 13.87
4 ao lucapa lucapa,ao -8.42 -8.42 64.81 77.0 0.0 1.70
5 ao namibe namibe,ao -15.19 -15.19 71.83 100.0 0.0 2.15

City Latitude Vs Max Temperature (F)

df_unique_cities.plot(kind="scatter",x="Lat",y="Temperature(F)",facecolor="b",edgecolor='black')
plt.title("City Latitude Vs Maximum Temperature (May26, 2018)")
plt.xlabel("Latitude");plt.ylabel("Max Temperature (F)")
plt.grid(linestyle='-.',color='black',linewidth=.5,alpha=.25)
plt.xlim(-60,80);plt.ylim(0,120)
plt.savefig("Output/latvstemp.png")

png

City Latitude Vs Humidity Plot

df_unique_cities.plot(kind="scatter",x="Lat",y="Humidity(%)",facecolor="b",edgecolor='black')
plt.title("City Latitude Vs Humidity (%) (May26, 2018)")
plt.xlabel("Latitude");plt.ylabel("Humidity (%)")
plt.grid(linestyle='-.',color='black',linewidth=.5,alpha=.25)
plt.xlim(-60,80);plt.ylim(0,120)
plt.savefig("Output/latvshumidity.png")

png

City Latitude Vs Cloudiness Plot

df_unique_cities.plot(kind="scatter",x="Lat",y="Cloudiness(%)",facecolor="b",edgecolor='black')
plt.title("City Latitude Vs Cloudiness (%) (May26, 2018)")
plt.xlabel("Latitude");plt.ylabel("Cloudiness (%)")
plt.grid(linestyle='-.',color='black',linewidth=.5,alpha=.25)
plt.xlim(-60,80);plt.ylim(0,120)
plt.savefig("Output/latvscloudiness.png")

png

City Latitude Vs Wind Speed Plot

df_unique_cities.plot(kind="scatter",x="Lat",y="Wind Speed(mph)",facecolor="b",edgecolor='black')
plt.title("City Latitude Vs Wind Speed (mph) (May26, 2018)")
plt.xlabel("Latitude");plt.ylabel("Wind Speed (mph)")
plt.grid(linestyle='-.',color='black',linewidth=.5,alpha=.25)
plt.xlim(-60,80);plt.ylim(0,45)
plt.savefig("Output/latvswind.png")

png

Conclusions

1. The cities near equator (0 latitude) have higher maximum temperature, with temperature falling down as the cities go farther from equator.
2. Humidity and cloudiness show no correlation to the latitudes. Here the data is quite evenly distributed.
3. Wind speed in most cities are less than 15mph which shows good mild weather today for most places. However some places have wind speed in higher range (15-25mph). In general we do not see any coorelation between wind speed and the latitude.