Study of COVID-19 Vaccine Hesitancy

Table of Contents

• 1  Introduction
• 2  Data and Interpretation
• 3  Finding
• 4  Future work
• 5  Appendix
  • 5.1  Load data
  • 5.2  User-defined functions
  • 5.3  Comparison between Categories (Distributed vs Administered)
    • 5.3.1  Grand comparison of Times series by Locations, measured in Number of vaccine
    • 5.3.2  Grand comparison of Daily change by Locations, measured in Number of vaccine
    • 5.3.3  Comparison of Time series by Vaccines and Locations, measured in Number of vaccine
    • 5.3.4  Comparison of Daily change by Vaccines and Locations, measured in Number of vaccine
    • 5.3.5  Comparison of Time series by Age range and Locations, measured in Number of vaccine
    • 5.3.6  Comparison of Daily change by Age range and Locations, measured in Number of vaccine
  • 5.4  Comparison between Categories (Series_complete vs Administered_Dose1_Recip)
    • 5.4.1  Comparison of Time series by Age range and Locations, measured in Population
    • 5.4.2  Comparison of Daily change by Age range and Locations, measured in Population
    • 5.4.3  Comparison of Time series by Age range and Locations, measured in Population proportion
    • 5.4.4  Comparison of Daily change by Age range and Locations, measured in Population proportion
  • 5.5  Comparison between Vaccine
    • 5.5.1  Comparison of Time series by Age range and Locations, measured in Population
    • 5.5.2  Comparison of Daily Change by Age range and Locations, measured in Population

Introduction¶

This study focus on discovering the signs of vaccine hesitancy based upon the overall population. The area of intereste are the states of Connecticut (CT) and South Dakota (SD). Note that CT has approximately four times in populaton compare to SD.

With the given data, the time series and the daily change distribution of the following aspects were investigated:

• Comparison between the number of Distributed vs Administered vaccine: Do the supply and consumption of the vaccine equal?
  • by Location, measured in Number of vaccine
  • by Vaccines type and Locations, measured in Number of vaccine
  • by Age range and Locations, measured in Number of vaccine
• Comparison between Series_complete vs Administered_Dose1_Receipt: Is vaccine hesitancy of completion of series?
  • by Age range and Locations, measured in Population
  • by Age range and Locations, measured in Population proportion
• Comparison between Vaccine: Is vaccine preferences exist?
  • by Age range and Locations, measured in Population

Data and Interpretation¶

Data	Interpretation
times series	time period from 1/22/2020 to 10/22/2021. were used to assess the overall trend and amount.
distribution of daily changes	calculated by taking the difference between previous day and the next day value. Used to assess the supply and demand corresponding to the distributed and administered dose. The distribution of daily changes count probably follow Poisson distribution or quasi-Poisson distribution, because the daily changes account for the number of events occurring in a fixed interval of time

Finding¶

• The supply and demand were increasing but not equilibrium.
  • The number of distributed and administered dose were both incrasing. (Section 5.3.1)
  • There is sigificant difference between the number of distributed and administered Dose for both states. (Section 5.3.2)
• In general, supply > demand
  • The ranking of the number of distributed and administered dose: Pfizer > Moderna > Janssen > Unknown manufacturer. The number of distributed dose were generally more than that of Administered dose. (Section 5.3.3)
  • Focus on Monderna and Pfizer, there is obvious difference betweeen the distribution of daily change of distributed and administered dose in both states. (Section 5.3.4)
• Excessive supply than demand in the 65+ group
  • The amount of distributed dose were always greater than the admistered dose for different age range and States. Noticeably, distributed dose is about 2~3 times of the administered dose for the 65+ population. (Section 5.3.5)
  • In general, the daily increase were higher in the number Distributed than Administered dose. Noticeably, the 65+ population has much daily increase of dose for Distributed than Administered. (Section 5.3.6)
• Vaccine hesitancy except 12+ population
  • The population of Series_complete was lagging behind the population Administered_Dose1_Recip, which match the common sentense. (Section 5.4.1)
  • The daily change distributions have Bimodal pattern (two peaks) for both Series_complete and Administered_Dose1_Recip. Excpet the 12+ population, the distribution of Administered_Dose1_Recip has more right-skewed but lower right peak than that of Series_complete. It implied that people were runing for the first dose, but not so demanding to complete the series. (Section 5.4.2)
• Mass Vaccination Event help people geting the frist dose
  • The propotion increase with approximately following the S-shaped growth curve (sigmoid growth curve) that positive acceleration phase (increases rapidly approaching an exponential growth rate) and negative acceleration phase (slowing down increases the until at zero growth rate the population stabilizes) (Section 5.4.3)
  • The high frequancy of small daily proportion increase in Series_complete might suggest that spontaneous behavior of vaccination for series completion. In contrast, the longer tail in the daily proportion increase distribution of Series_complete refer to the contribution by the Mass Vaccination Event. (Section 5.4.4)
• Vaccine hesitancy for Janssen
  • the accumulative dose consumption for Pfizer and Moderna follow exponential growth curve, whereas that for Janssen seem start with exponential growth curve but later change to a linear curve with slower rate than other vaccine. The change in growth pattern probably cause by the reveal of the risk for a rare but serious adverse event—blood clots with low platelets after receiving the J&J/Janssen COVID-19 Vaccine. (Section 5.5.1)
  • There is no difference in daily change distribution for Pfizer and Moderna, but that for Janssen has smaller peak and shorter tail, implying smaller demand for Janssen. (Section 5.5.2)

Future work¶

• Perform two-sample Kolmogorov-Smirnov test for goodness of fit to make comparison
• Calculate the ratio between daily rate of distributed and Administered: $\frac{\delta \text{distributed dose}}{\delta \text{Administered dose}}$ to quantify the spontaneous demand/supply and assess the distributing efficiency.

Appendix¶

Load data¶

In [1]:

import copy 
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from scipy import stats

%matplotlib notebook
%matplotlib inline

In [2]:

vacc = pd.read_excel('raw data.xlsx', sheet_name='COVID-19_Vaccinations_in_the_Un')
case = pd.read_excel('raw data.xlsx', sheet_name='United_States_COVID-19_Cases_an')

In [3]:

vacc = vacc[(vacc["Location"] == 'CT') | (vacc["Location"] == 'SD')]
vacc=vacc.rename({'Dist_Per_100K': 'Distributed_Per_100K'}, axis=1)
vacc = vacc.rename({'Administered_Dose1_Recip': 'Administered_Dose1_Recip_Yes'}, axis=1)
vacc = vacc.rename({'Administered_Dose1_Pop_Pct': 'Administered_Dose1_Recip_Pop_Pct'}, axis = 1)

In [4]:

print(f'COVID-19_Vaccinations_in_the_Un: \n {vacc.columns}')
print(f'United_States_COVID-19_Cases_an: \n {case.columns}')

COVID-19_Vaccinations_in_the_Un: 
 Index(['Date', 'MMWR_week', 'Location', 'Distributed', 'Distributed_Janssen',
       'Distributed_Moderna', 'Distributed_Pfizer', 'Distributed_Unk_Manuf',
       'Distributed_Per_100K', 'Distributed_Per_100k_12Plus',
       'Distributed_Per_100k_18Plus', 'Distributed_Per_100k_65Plus',
       'Administered', 'Administered_12Plus', 'Administered_18Plus',
       'Administered_65Plus', 'Administered_Janssen', 'Administered_Moderna',
       'Administered_Pfizer', 'Administered_Unk_Manuf', 'Admin_Per_100K',
       'Admin_Per_100k_12Plus', 'Admin_Per_100k_18Plus',
       'Admin_Per_100k_65Plus', 'Recip_Administered',
       'Administered_Dose1_Recip_Yes', 'Administered_Dose1_Recip_Pop_Pct',
       'Administered_Dose1_Recip_12Plus',
       'Administered_Dose1_Recip_12PlusPop_Pct',
       'Administered_Dose1_Recip_18Plus',
       'Administered_Dose1_Recip_18PlusPop_Pct',
       'Administered_Dose1_Recip_65Plus',
       'Administered_Dose1_Recip_65PlusPop_Pct', 'Series_Complete_Yes',
       'Series_Complete_Pop_Pct', 'Series_Complete_12Plus',
       'Series_Complete_12PlusPop_Pct', 'Series_Complete_18Plus',
       'Series_Complete_18PlusPop_Pct', 'Series_Complete_65Plus',
       'Series_Complete_65PlusPop_Pct', 'Series_Complete_Janssen',
       'Series_Complete_Moderna', 'Series_Complete_Pfizer',
       'Series_Complete_Unk_Manuf', 'Series_Complete_Janssen_12Plus',
       'Series_Complete_Moderna_12Plus', 'Series_Complete_Pfizer_12Plus',
       'Series_Complete_Unk_Manuf_12Plus', 'Series_Complete_Janssen_18Plus',
       'Series_Complete_Moderna_18Plus', 'Series_Complete_Pfizer_18Plus',
       'Series_Complete_Unk_Manuf_18Plus', 'Series_Complete_Janssen_65Plus',
       'Series_Complete_Moderna_65Plus', 'Series_Complete_Pfizer_65Plus',
       'Series_Complete_Unk_Manuf_65Plus', 'Additional_Doses',
       'Additional_Doses_Vax_Pct', 'Additional_Doses_18Plus',
       'Additional_Doses_18Plus_Vax_Pct', 'Additional_Doses_50Plus',
       'Additional_Doses_50Plus_Vax_Pct', 'Additional_Doses_65Plus',
       'Additional_Doses_65Plus_Vax_Pct', 'Additional_Doses_Moderna',
       'Additional_Doses_Pfizer', 'Additional_Doses_Janssen',
       'Additional_Doses_Unk_Manuf'],
      dtype='object')
United_States_COVID-19_Cases_an: 
 Index(['submission_date', 'state', 'tot_cases', 'conf_cases', 'prob_cases',
       'new_case', 'pnew_case', 'tot_death', 'conf_death', 'prob_death',
       'new_death', 'pnew_death', 'created_at', 'consent_cases',
       'consent_deaths'],
      dtype='object')

User-defined functions¶

In [5]:

def replace_leading_ts(ts):   
    """
    replace the leading zero in times series with nan, in order to be removed
    """
    ts.loc[ts.eq(0).cumprod().astype(bool)]= np.float64('nan')
    return ts

In [6]:

def diff_bygroup(df, sorting_col_list, date = ['Date']):
    """
    calculate the daily change by groups
    """
    df2 = df.sort_values(sorting_col_list)
    df2['value'] = df2.groupby(sorting_col_list)['value'].transform(lambda x: replace_leading_ts(x))
    df2['diffs'] = df2.groupby(sorting_col_list)['value'].transform(lambda x: x.diff(periods=1)) 

    df2 = df2.sort_index()
    return df2

Comparison between Categories (Distributed vs Administered)¶

Grand comparison of Times series by Locations, measured in Number of vaccine¶

-'Distributed', -'Administered'

In [7]:

# data cleaning
Grand_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                           value_vars=['Distributed', 'Administered'], 
                           ignore_index=False)
Grand_long = Grand_long.rename({'variable': 'Category'}, axis=1)
Grand_long = Grand_long.reset_index(drop=True)

In [8]:

# Visualization
sns.set(font_scale=1)  
g = sns.FacetGrid(Grand_long, col = 'Location', size=5,
                  legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.8) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Grand comparison of Times series by Locations, measured in Number of vaccine')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Number of vaccine")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[8]:

<seaborn.axisgrid.FacetGrid at 0x13f300eb0>

Grand comparison of Daily change by Locations, measured in Number of vaccine¶

In [9]:

# data cleaning
Grand_long_diff = diff_bygroup(df = Grand_long, 
             sorting_col_list=['Location', 'Category'])
Grand_long_diff = Grand_long_diff.dropna(axis = 0)

In [10]:

# data cleaning
Grand_long_diff = diff_bygroup(df = Grand_long, 
             sorting_col_list=['Location', 'Category'])
Grand_long_diff = Grand_long_diff.dropna(axis = 0)

# Visualization
sns.set(font_scale=1.5)  
g = sns.FacetGrid(Grand_long_diff, col = 'Location', hue = 'Category', size=7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.8) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Grand comparison of Daily change by Locations, measured in Number of vaccine')
g.set_axis_labels("Daily change of Number of vaccine", 'Count')
g.add_legend()
g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[10]:

<seaborn.axisgrid.FacetGrid at 0x13fef7fa0>

In [11]:

# two-sample Kolmogorov-Smirnov test for goodness of fit: 
# p-value for testing if distribution differ by categoris. 
# tests are run for each subplot.
Grand_long_diff.drop(columns=['value']).\
    pivot(index=['Date', 'Location'], columns='Category', values='diffs').\
    reset_index().\
    drop(columns=['Date']).\
    groupby(['Location']).\
    apply(lambda df: stats.ks_2samp(df.Administered, df.Distributed)[1])

Out[11]:

Location
CT    3.013765e-12
SD    4.483280e-18
dtype: float64

Comparison of Time series by Vaccines and Locations, measured in Number of vaccine¶

• 'Distributed_Janssen', 'Distributed_Moderna', 'Distributed_Pfizer', 'Distributed_Unk_Manuf'
• 'Administered_Janssen', 'Administered_Moderna', 'Administered_Pfizer', 'Administered_Unk_Manuf'

In [12]:

# data cleaning
Vaccines_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                           value_vars=['Distributed_Janssen', 'Distributed_Moderna', 'Distributed_Pfizer', 'Distributed_Unk_Manuf',
                                       'Administered_Janssen', 'Administered_Moderna', 'Administered_Pfizer', 'Administered_Unk_Manuf'], 
                           ignore_index=False)

Category_Vaccine = Vaccines_long.variable.str.split('_', n = 1, expand = True)
Category_Vaccine = Category_Vaccine.rename({0: 'Category', 1: 'Vaccine'}, axis=1)  
Vaccines_long = pd.concat([Vaccines_long, Category_Vaccine], axis=1)
Vaccines_long = Vaccines_long.reset_index(drop=True)

In [13]:

# Visualization
sns.set(font_scale=2) 
g = sns.FacetGrid(Vaccines_long, col='Vaccine', row = 'Location', size = 7,
                  legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.85) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Comparison of Time series by Vaccines and Locations, measured in Number of vaccine')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Number of vaccine")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[13]:

<seaborn.axisgrid.FacetGrid at 0x141204f40>

Comparison of Daily change by Vaccines and Locations, measured in Number of vaccine¶

In [14]:

# data cleaning
Vaccines_long_diff = diff_bygroup(df = Vaccines_long, 
             sorting_col_list=['Location', 'Category' , 'Vaccine'])
Vaccines_long_diff = Vaccines_long_diff.dropna(axis = 0)

In [15]:

# Visualization
sns.set(font_scale=1.5)  
g = sns.FacetGrid(Vaccines_long_diff, col='Vaccine', row = 'Location', hue = 'Category', size =7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.85) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Comparison of Daily change by Vaccines and Locations, measured in Number of vaccine')
g.set_axis_labels("Daily change of Number of vaccine", "Count")
g.add_legend()
g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[15]:

<seaborn.axisgrid.FacetGrid at 0x13f3eadf0>

Comparison of Time series by Age range and Locations, measured in Number of vaccine¶

• 'Distributed_Per_100K', 'Distributed_Per_100k_12Plus','Distributed_Per_100k_18Plus', 'Distributed_Per_100k_65Plus',
• 'Admin_Per_100K', 'Admin_Per_100k_12Plus', 'Admin_Per_100k_18Plus', 'Admin_Per_100k_65Plus',

In [16]:

# data cleaning
Age_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                           value_vars=['Distributed_Per_100K', 'Distributed_Per_100k_12Plus','Distributed_Per_100k_18Plus', 'Distributed_Per_100k_65Plus',
                                       'Admin_Per_100K', 'Admin_Per_100k_12Plus', 'Admin_Per_100k_18Plus', 'Admin_Per_100k_65Plus'], 
                           ignore_index=False)
Category_Per_100K_Age = Age_long.variable.str.split('_', n = 1, expand = True)
Category_Per_100K_Age = Category_Per_100K_Age.rename({0: 'Category', 1: 'Per_100K_Age'}, axis=1)  
Age_long = pd.concat([Age_long, Category_Per_100K_Age], axis=1)
Age_long = Age_long.reset_index(drop=True)

In [17]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Age_long, col='Per_100K_Age', row = 'Location', size = 7,
                  legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Comparison of Time series by Age range and Locations, measured in Number of vaccine')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Number of vaccine")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[17]:

<seaborn.axisgrid.FacetGrid at 0x1418a2b80>

Comparison of Daily change by Age range and Locations, measured in Number of vaccine¶

In [18]:

# data cleaning
Age_long_diff = diff_bygroup(df = Age_long, 
             sorting_col_list=['Location', 'Category' , 'Per_100K_Age'])
Age_long_diff = Age_long_diff.dropna(axis = 0)

# Visualization
sns.set(font_scale=1.5)  
g = sns.FacetGrid(Age_long_diff, col='Per_100K_Age', row = 'Location', hue = 'Category', size = 7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.85) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Distributed vs Administered): \n Comparison of Daily change by Age range and Locations, measured in Number of vaccine')
g.set_axis_labels("Daily change of number of vaccine", 'count')
g.add_legend()
g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[18]:

<seaborn.axisgrid.FacetGrid at 0x140ca9520>

Comparison between Categories (Series_complete vs Administered_Dose1_Recip)¶

Comparison of Time series by Age range and Locations, measured in Population¶

• 'Administered_Dose1_Recip', 'Administered_Dose1_Recip_12Plus', 'Administered_Dose1_Recip_18Plus', 'Administered_Dose1_Recip_65Plus',

• 'Series_Complete_Yes', 'Series_Complete_12Plus', 'Series_Complete_18Plus', 'Series_Complete_65Plus'

In [19]:

# data cleaning
Dose_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                   value_vars=['Administered_Dose1_Recip_Yes', 
                               'Administered_Dose1_Recip_12Plus',
                               'Administered_Dose1_Recip_18Plus',
                               'Administered_Dose1_Recip_65Plus',
                               'Series_Complete_Yes',
                               'Series_Complete_12Plus',
                               'Series_Complete_18Plus',
                               'Series_Complete_65Plus'], 
                           ignore_index=False)
Category_Per_100K_Age = Dose_long.variable.str.rsplit('_', n = 1, expand = True)
Category_Per_100K_Age = Category_Per_100K_Age.rename({0: 'Category', 1: 'Age'}, axis=1)  
Dose_long = pd.concat([Dose_long, Category_Per_100K_Age], axis=1)
Dose_long = Dose_long.reset_index(drop=True)

In [20]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Dose_long, col='Age', row = 'Location', size = 7,
                  legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Series_complete vs Administered_Dose1_Recip): \n Comparison of Time series by Age range and Locations, measured in Population')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Population")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[20]:

<seaborn.axisgrid.FacetGrid at 0x141ca73d0>

Comparison of Daily change by Age range and Locations, measured in Population¶

In [21]:

# data cleaning
Dose_long_diff = diff_bygroup(df = Dose_long, 
             sorting_col_list=['Location', 'Category', 'Age'])
Dose_long_diff = Dose_long_diff.dropna(axis = 0)

In [22]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Dose_long_diff, col='Age', row = 'Location', hue = 'Category', size = 7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Series_complete vs Administered_Dose1_Recip): \n Comparison of Daily change by Age range and Locations, measured in Population')
g.set_axis_labels("Daily change of population", 'count')
g.add_legend()
g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[22]:

<seaborn.axisgrid.FacetGrid at 0x142b01af0>

Comparison of Time series by Age range and Locations, measured in Population proportion¶

• 'Administered_Dose1_Recip_Pop_Pct', 'Administered_Dose1_Recip_12PlusPop_Pct', 'Administered_Dose1_Recip_18PlusPop_Pct', 'Administered_Dose1_Recip_65PlusPop_Pct',

• 'Series_Complete_Pop_Pct', 'Series_Complete_12PlusPop_Pct', 'Series_Complete_18PlusPop_Pct', 'Series_Complete_65PlusPop_Pct',

In [23]:

# data cleaning
Pop_Pct_Age_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                    value_vars=['Administered_Dose1_Recip_Pop_Pct',
                                'Administered_Dose1_Recip_12PlusPop_Pct',
                                'Administered_Dose1_Recip_18PlusPop_Pct',
                                'Administered_Dose1_Recip_65PlusPop_Pct', 
                                'Series_Complete_Pop_Pct', 
                                'Series_Complete_12PlusPop_Pct', 
                                'Series_Complete_18PlusPop_Pct', 
                                'Series_Complete_65PlusPop_Pct',], 
                           ignore_index=False)
Pop_Pct_Age_long['Category'] = Pop_Pct_Age_long.variable.str.extract('(^Administered_Dose1_Recip|^Series_Complete)')
Pop_Pct_Age_long['Pop_Pct_Age'] = Pop_Pct_Age_long.variable.str.extract('(Pop_Pct|12PlusPop_Pct|18PlusPop_Pct|65PlusPop_Pct)$')
Pop_Pct_Age_long = Pop_Pct_Age_long.reset_index(drop=True)

In [24]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Pop_Pct_Age_long, col='Pop_Pct_Age', row = 'Location',  size = 7, legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Series_complete vs Administered_Dose1_Recip): \n Comparison of Time series by Age range and Locations, measured in Population proportion')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Population proportion, %")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[24]:

<seaborn.axisgrid.FacetGrid at 0x143c39820>

Comparison of Daily change by Age range and Locations, measured in Population proportion¶

In [25]:

# data cleaning
Pop_Pct_Age_long_diff = diff_bygroup(df = Pop_Pct_Age_long, 
             sorting_col_list=['Location', 'Category', 'Pop_Pct_Age'])
Pop_Pct_Age_long_diff = Pop_Pct_Age_long_diff.dropna(axis = 0)

In [26]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Pop_Pct_Age_long_diff, col='Pop_Pct_Age', row = 'Location', hue = 'Category', size = 7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.kdeplot, 'diffs')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Series_complete vs Administered_Dose1_Recip): \n Comparison of Daily change by Age range and Locations, measured in Population proportion')
g.set_axis_labels("Daily change of Population proportion", "density")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[26]:

<seaborn.axisgrid.FacetGrid at 0x144040250>

In [27]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Pop_Pct_Age_long_diff, col='Pop_Pct_Age', row = 'Location', hue = 'Category', size = 7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Categories (Series_complete vs Administered_Dose1_Recip): \n Comparison of Daily change by Age range and Locations, measured in Population proportion')
g.set_axis_labels("Daily change of Population proportion", "Count")
g.add_legend()
# g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[27]:

<seaborn.axisgrid.FacetGrid at 0x13f338250>

Comparison between Vaccine¶

Comparison of Time series by Age range and Locations, measured in Population¶

'Series_Complete_Janssen', 'Series_Complete_Moderna', 'Series_Complete_Pfizer', 'Series_Complete_Unk_Manuf', 'Series_Complete_Janssen_12Plus', 'Series_Complete_Moderna_12Plus', 'Series_Complete_Pfizer_12Plus', 'Series_Complete_Unk_Manuf_12Plus', 'Series_Complete_Janssen_18Plus', 'Series_Complete_Moderna_18Plus', 'Series_Complete_Pfizer_18Plus', 'Series_Complete_Unk_Manuf_18Plus', 'Series_Complete_Janssen_65Plus', 'Series_Complete_Moderna_65Plus', 'Series_Complete_Pfizer_65Plus', 'Series_Complete_Unk_Manuf_65Plus'

In [28]:

# data cleaning
Series_Complete_long = pd.melt(vacc, id_vars=['Date', 'Location'], 
                           value_vars=['Series_Complete_Janssen',
                                       'Series_Complete_Moderna', 
                                       'Series_Complete_Pfizer',
                                       'Series_Complete_Unk_Manuf', 
                                       'Series_Complete_Janssen_12Plus',
                                       'Series_Complete_Moderna_12Plus', 
                                       'Series_Complete_Pfizer_12Plus',
                                       'Series_Complete_Unk_Manuf_12Plus', 
                                       'Series_Complete_Janssen_18Plus',
                                       'Series_Complete_Moderna_18Plus', 
                                       'Series_Complete_Pfizer_18Plus',
                                       'Series_Complete_Unk_Manuf_18Plus', 
                                       'Series_Complete_Janssen_65Plus',
                                       'Series_Complete_Moderna_65Plus', 
                                       'Series_Complete_Pfizer_65Plus',
                                       'Series_Complete_Unk_Manuf_65Plus'], 
                           ignore_index=False)
Series_Complete_long['Category'] = Series_Complete_long.variable.str.extract('(Janssen|Moderna|Pfizer|Unk_Manuf)')
Series_Complete_long['Pop_Pct_Age'] = Series_Complete_long.variable.str.extract('(12Plus|18Plus|65Plus)$')
Series_Complete_long = Series_Complete_long.reset_index(drop=True)

In [29]:

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Series_Complete_long, col='Pop_Pct_Age', row = 'Location', size = 7, legend_out=True)
g.map_dataframe(sns.lineplot, x= 'Date', y= 'value', hue = 'Category')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Vaccines: \n Comparison of Time series by Age range and Locations, measured in Population')
g.set_xticklabels(rotation = 30)
g.set_axis_labels("Date", "Population")
g.add_legend()

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[29]:

<seaborn.axisgrid.FacetGrid at 0x146590430>

Comparison of Daily Change by Age range and Locations, measured in Population¶

In [30]:

# data cleaning
Series_Complete_long_diff = diff_bygroup(df = Series_Complete_long, 
             sorting_col_list=['Location', 'Category', 'Pop_Pct_Age'])
Series_Complete_long_diff = Series_Complete_long_diff.dropna(axis = 0)

# Visualization
sns.set(font_scale=1.5) 
g = sns.FacetGrid(Series_Complete_long_diff, col='Pop_Pct_Age', hue = 'Location',  row= 'Category', size = 7,
                  palette=sns.color_palette('bright')[:3],
                  legend_out=True)
g.map(sns.histplot, 'diffs')

g.fig.subplots_adjust(top=0.9) # adjust the Figure in rp
g.fig.suptitle('Comparison between Vaccines: \n Comparison of Daily Change by Age range and Locations (CT vs SD), measured in Population')

g.set_axis_labels("Daily change in population", "Count")
g.add_legend()
g.set(yscale='log')

/usr/local/lib/python3.9/site-packages/seaborn/axisgrid.py:337: UserWarning: The `size` parameter has been renamed to `height`; please update your code.
  warnings.warn(msg, UserWarning)

Out[30]:

<seaborn.axisgrid.FacetGrid at 0x14634a130>

Converting to pdf without code:

% jupyter nbconvert --to pdf --TemplateExporter.exclude_input=True database_final_project.ipynb