Create a Program to Implement NYC Datasets in Python Assignment Solution.

# Datasets

# Datasets

###COVID-19 Daily Counts of Cases, Hospitalizations, and DeathsHealth

https://data.cityofnewyork.us/Health/COVID-19-Daily-Counts-of-Cases-Hospitalizations-an/rc75-m7u3

**Download link:** https://data.cityofnewyork.us/api/views/rc75-m7u3/rows.csv?accessType=DOWNLOAD

### Emergency Department Visits and Admissions for Influenza-like Illness and/or Pneumonia

https://data.cityofnewyork.us/Health/Emergency-Department-Visits-and-Admissions-for-Inf/2nwg-uqyg

**Download link:** https://data.cityofnewyork.us/api/views/2nwg-uqyg/rows.csv?accessType=DOWNLOAD

The datasets mentioned above have information on the number of cases of infection and death reported in NY per day. The other dataset contains the number of visits to hospitals or ER rooms for cases of pneumonia, influenza or other similar symptoms. It is planned to verify a direct relationship between these datasets from the date the first cases of COVID-19 were reported in NY.

import pandas as pd

import requests

import matplotlib.pyplot as plt

import io

import seaborn as sns

## Step 1: Download Datasets

data1 = pd.read_csv('https://data.cityofnewyork.us/api/views/rc75-m7u3/rows.csv?accessType=DOWNLOAD')

data1.head()

data2 = pd.read_csv('https://data.cityofnewyork.us/api/views/2nwg-uqyg/rows.csv?accessType=DOWNLOAD')

data2.head()

# Step 2: Clean Datasets

### For first dataset, select only the columns of interest

columns_of_interest = ['DATE_OF_INTEREST', 'CASE_COUNT', 'HOSPITALIZED_COUNT', 'DEATH_COUNT']

df1 = data1[columns_of_interest]

### Convert 'DATE_OF_INTEREST' to Datetime

df1['DATE_OF_INTEREST'] = pd.to_datetime(df1['DATE_OF_INTEREST'])

df1.head()

### Let's do the same for the second dataset

columns_of_interest= ['extract_date', 'date', 'total_ed_visits', 'ili_pne_visits']

df2 = data2[columns_of_interest]

df2['extract_date'] = pd.to_datetime(df2['extract_date'])

df2['date'] = pd.to_datetime(df2['date'])

df2.head()

# Let's find the date for the first COVID-19 case reported

start_date = data1.loc[0, 'DATE_OF_INTEREST']

print(start_date)

### Find all rows in second dataset from start_date to present

df2 = df2[df2['date'] >= start_date]

df2.head()

# Step 3: Plots

### Plot the number of COVID-19 case and number of ER visits per day

**NOTE: ** We will normalize (between 0 and 1) the number of cases for both datasets. This is because we only want to compare the shape of curves and not the values

df1_grouped = df1.groupby(by=['DATE_OF_INTEREST']).sum().sort_values(by=['DATE_OF_INTEREST'], ascending=True)

df1_grouped = (df1_grouped-df1_grouped.min())/(df1_grouped.max() - df1_grouped.min())

df1_grouped.head()

df2_grouped = df2.groupby(by=['date']).sum().sort_values(by=['date'], ascending = True)

df2_grouped = (df2_grouped-df2_grouped.min())/(df2_grouped.max() - df2_grouped.min())

df2_grouped.head()

### Plot

plt.figure()

ax = df1_grouped.plot(y = 'CASE_COUNT', label = 'COVID-19 Cases')

df2_grouped.plot(y='ili_pne_visits', label = 'Hospital Visits', ax = ax)

plt.legend()

plt.show()

We see that there is a clear correlation between the curves. At the beginning, the curves are very similar, and it is because when the pandemic began, everyone was very scared and at the first symptom (even if it was minimal) people attended the ER room. As time passed and social distancing measures began to be implemented, we see that the number of covid cases decreased as did visits to the ER. By October 2020, the number of cases increased again (second wave) and ER visits also increased, although in lesser quantity and it is because people were no longer so scared.

## Correlation maps for each dataset

corr1 = data1.corr()

corr2 = data2.corr()

### Dataset 1

sns.heatmap(corr1)

## Dataset 2

sns.heatmap(corr2)