Ibrahim Oladiti

FUNDAMENTALS OF DATA SCIENCE CENSUS DATA REPORT INTRODUCTION We are presented with the 1881 census record of a moderately sized town between two large cities. The town does not have a university but has university students who reside in the town and commute to the university in the neighboring city. The objective of this report is to examine the census data, employ the needed data cleaning and analysis methods and develop insights that will enable the local government make decision on • What to build on an unoccupied piece of land • What to invest in in the town To answer these questions, the report is divided into 3 sections. The first section is data cleaning which is done correct inappropriate entry or errors in our data. The second section is data analysis and visualization to get insight into the attribute of the population. The last section of the report discusses the insight and provides recommendations to accomplish our objectives. DATA CLEANING To clean the data, first the data is checked for duplicate records, then the columns that make up the data frame is examined to know the columns that is important for analysis. After loading the census csv dataset to the Pandas data frame, it had 12 columns. The first column named ‘unnamed 0:’ which contains the serial numbers of the records was dropped from the data frame because it is not important to our analysis. The data was cleaned by looking at the unique values of each column to see if the response was correctly imported. On the House Number column, incorrect input ‘Four’ was discovered. This was replaced with the correct input of ‘4’ and the column was converted to the integer datatype. On the First Name and Surname columns, it was discovered that one record had a missing first name and four records had a missing surname. There was no record with both first name and surname missing. These records are left on the data frame because the surname and first name are not needed for our analysis. The Age column had some incorrect input like ’-’ and ‘eight’. This was corrected by replacing it with the integer 8 and converting float values to integer. Three records were found to have no input for age. This was corrected by replacing the empty string with the mode of the Age column. Two records were missing input in the Relationship to Head of House column. This was corrected by looking at their individual address to see the properties of their family and age to determine the most appropriate relationship to be assigned. On the Marital Status and Gender columns inputs where abbreviations were used are corrected to their correct values examples are ‘F’ for Female and ‘M’ for male in the Gender column and ‘M’ for Married and ‘D’ for Divorced on the Marital Status column. Marital Status for minors was inputted as NA meaning ‘Not Applicable’. The legal marriage age is 16 years for female with the consent of the parent while that of male is 18 years old. A household where the Head is a 17-year-old student, and all other occupants are lodgers. The record for the household was dropped from the dataset because a responded must be 18 years and above to be considered the Head of Household (Sharfman, 2022). Outliers was removed for record that showed a 35-year-old man with occupation as ‘Child’, this is dropped from the data frame. On the religion column, the religion for minors is recorded as NA this is seen as NaN by pandas, so it was replaced with string ‘NA’ for it to considered during analysis. Two records with ‘Sith’ religion were dropped from the data frame. This is because ‘Sith’ is not a real-world religion, it is used as a fictional religion in the move franchise ‘Star Wars’ (Wikipedia, 2022). There are two records with the input ‘Undecided’ and one with ‘Private’ for religion. Since the number of records with this input is not statistically significant, the input is replaced with ‘None’ to get a cleaner analysis. ANALYSIS AND INSIGHTS Population Demographics After data cleaning, the initial insight into the population is done to get a basic understanding of the populations and its unique attributes. Insights from the population demography will inform on attributes that need further analysis. Population pyramid The population pyramid is a graphical representation of the distribution of age group by gender. To plot the age pyramid, a new column for age-group was added to the data frame. This is done by grouping the age in five internals. Analysis of the age distribution shows that the town does not have an aging population a considerable number of people are of working age which is between 16-64 years (GOV.UK, 2018). Since the median age of the town is 35, it means that there is no pressing need for the construction of care homes for the elderly and more opportunities should be provided for mostly youthful populations. Figure 1: Population Pyramid of the town residents Marital Status The legal age for marriage in the United Kingdom is 18 years (GOV.UK, 2022). Marital status for people under 18 years is inputted as Not Applicable. Minors represent 25% of the total population of the town. Analysis of those that are eligible for marriage shows that the majority (45%) are single, 38% are married, 12% are divorced and 4.8% responded as widowed. Approximately 50% of the divorced people in the town are female, which mean that most men leave the town after divorced. The slightly higher percentage of female population in the town is normal considering that there is more female population in the town. Figure 2: Marital status by population Figure 3: Marital Status by Age Religion Religion for minors (people less than 18 years) is recorded as ‘NA’ meaning not applicable. 45% of the others responded to having no religion, 28% responded that they are Christians, 14% said they were Catholics, 9% responded as Methodists, while the other religions represent 3% of the total population. Figure 4: Religion Distribution Analysis of the religion shows that although there are more Christian than any other religion, majority of the residents responded to not having any religion. The chart also show that Christians have a growing population in the town. While the construction of a Christian church might be needed in the future, it is not much of a high priority now given that it will not benefit a sizeable percentage of the population. Figure 5: Religion by Age Infirmity 99.2% of the entire population responded to having no infirmity, while only 0.8% responded to having one of the six infirmities recorded. The population of those with infirmity is not significant enough to dedicate any investment in infirmity. Occupation and Unemployment To better analyze the occupation of the town inhabitants, the occupation field is further classified into six distinct groups. This groups are: • Child: These people to respond with ‘Child’ as occupation. They are between the ages 0-4 years and represent 5.8% of the total population. • Student: Responses with ‘Student’ as occupation is left as a group on its own. They are between the ages of 5 – 18 years and are 19.5% of the population. • University Student: These are the university student in the town. PhD students are also classified as university students. They are between the ages of 19 – 32 years. University students are 6.9% of the total population. • Unemployed: 5.9% of the population reported to be unemployed. 71.5% of the unemployed population are between the age 30-55 years. • Employed: 53.5% of the entire population reported to in a form of employment. This is evenly distributed across the entire working age group (16-64 years) • Retired: These are people in the town that are retired they are 65 years and above and represent 8.2% of the population Figure 6: Occupation Age Distribution Unemployment Statistics 5.9% of the population responded as unemployed in the population field. This only give us an overview of the unemployment in the town. To get a detailed analysis of analysis of the unemployment situation, we would have to use percentage unemployment and unemployment rate. The percentage unemployment and unemployment rate of the town is calculated with the formula below. 𝑇𝑜𝑡𝑎𝑙 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒 𝑜𝑓 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑒𝑑 = 𝑈𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑚𝑒𝑛𝑡 𝑟𝑎𝑡𝑒 = 𝑈𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑒𝑑 𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙 𝑊𝑜𝑟𝑘𝑖𝑛𝑔 𝑎𝑔𝑒 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑢𝑛𝑒𝑚𝑝𝑙𝑜𝑦𝑒𝑑 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑖𝑛 𝑙𝑎𝑏𝑜𝑢𝑟 𝑓𝑜𝑟𝑐𝑒 × 100 (Lee, 2021) × 100 (Lee, 2021) 5,647 people in the town reported to be engaged in one form of employment. 625 individuals reported to be unemployed, and the total number of working age population in the town is 7,218. Using the above formular, the percentage unemployment in the town is 8.7% while the unemployment rate is 11%. This is more than what is considered a healthy or acceptable unemployment rate 4-5% (Kim, 2022). This percentage indicates that investment in skills acquisition and training should be strongly considered in the town. There is more female unemployed than male in the town, so any investment in skills empowerment and trainings for the unemployed should put into consideration the type that is favourable with females. Figure 7:Unemployed Age Distribution Regular Commuters To assess the type of investment that should be made in the town, it is important to analyze the number of commuters in the town. Lodgers are people who rent from live with the homeowner. 447 people indicated to be lodgers in the town. The occupation of the lodgers are university students and people who work in either of the two adjacent cities close to the town and who like to benefit from the low rent in the town. There are 695 University students (including PhD Students) who are not lodgers but reside with their families or rent in a house not shared with a family and will also need to commute to the university. 5,276 people in the town are employed and are not lodgers. These are permanent residence in the town or rent houses not shared with a family. Looking at the type of occupation of the employed, it shows that there is a higher percentage of specialists and professionals living in the town who will require to larger cities for work. 25% (1,319) of this population is assumed to be employed outside of the town and are commuters. To get the total number of commuters in the town, we use the formula below. 𝑇𝑜𝑡𝑎𝑙 𝑛𝑢𝑚𝑏𝑒𝑟 𝑟𝑒𝑔𝑢𝑙𝑎𝑟 𝑜𝑓 𝑐𝑜𝑚𝑚𝑢𝑡𝑒𝑟𝑠 = 𝐿𝐷 + 𝑈𝑆 + 𝑂𝐸 Where LD = Number of lodgers US = Number University Students OE = Other Employed Commuters Using the formula above, the number of regular commuters in the town is calculated to be 2,461. Which is approximately 23% of the town’s population. This show that there is a considerable number of commuters to justify the construction of railway system in the town. Household occupancy Analysis of the relationships within household show that children tend to live with their parents well into adulthood and some household have three generations living together. This could be the house is big enough to accommodate an extended family indicating affluence of the household. Figure 8: Relationship to Head of House by Age There are 3,546 households in the town with an average household occupancy of 3 with most houses having only 2 occupants. This is a modest figure considering the average number of bedrooms in a house is 3.2 (Statista, 2018). This shows that there is no pressing need for investment in housing. Although this can be considered for future investment. Birth Rate and Death Rate There are 100 new births in the town. This is gotten from the number of residents that is less than one year. To calculate the birth and death rate the formula below is used., 𝐵𝑖𝑟𝑡ℎ 𝑟𝑎𝑡𝑒 = 𝑏𝑖𝑟𝑡ℎ𝑠 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 × 1000 (Watt-Watchers, 2022) 𝐷𝑒𝑎𝑡ℎ𝑠 𝐷𝑒𝑎𝑡ℎ 𝑟𝑎𝑡𝑒 = 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 × 1000 (Watt-Watchers, 2022) Using the formular above the birth rate for the town is 9.5 births per 1000 of the population. The death rate for the population could not be calculated because the number of deaths was not recorded in the census. Analysing of the population pyramid above, it could be deduced that there will be slight increase in population. RECOMMENDATIONS AND CONCLUSION Analysis of the census data showed that the town have a large population of people of working age that are mostly students and employed individuals. But there is still a high employment rate of 11%. 70% of the unemployed are between 30-50 years, indicating that they are most probably individual with skills that are no longer in high demand. To remedy this, it is recommended that skills acquisition centre be built on the unoccupied piece of land. The centre will train unemployed individuals if skills that are in high demand. Given the number of lodgers, and university students in the town, it is estimated 23% of the population will commute regularly to adjacent cities. Included in this estimate is 25% of employed population that is assumed to commute to work in the larger city. The number of commuters justifies investment in rail network because a significant percentage of the population will benefit from it. This will result in further increase in the population size in the nearest future. The study is limited by available data which does not include records of death which is needed to appropriately calculate the death rate. The records for income were not captured in the data collection. This would have allowed us to compute the disposable income and know the affluence of the population, which is a good metric needed to understand the economy of the town. A broader insight would have been achieved if a multi-year population data is provided to see progression of the population metrics. References GOV.UK, 2018. Working age population. [Online] Available at: https://www.ethnicity-facts-figures.service.gov.uk/uk-population-byethnicity/demographics/working-age-population/latest [Accessed 04 December 2022]. GOV.UK, 2022. GOV.UK. [Online] Available at: https://www.gov.uk/government/news/implementation-of-the-marriage-and-civilpartnership-minimum-age-act2022#:~:text=The%20Marriage%20and%20Civil%20Partnership%20(Minimum%20Age)%20Act%202 022%20received,on%20Monday%2027%20February%202023.&text=The%20Act%20 [Accessed 04 December 2022]. HackettT, I. W. C. a. C., 2022. Pew Research Center. [Online] Available at: https://www.pewresearch.org/fact-tank/2022/08/31/global-population-skews-malebut-un-projects-parity-between-sexes-by2050/#:~:text=In%202021%2C%20the%20global%20sex,and%20following%20pregnancy%20and%20 childbirth. [Accessed 04 December 2022]. Kim, P., 2022. The unemployment rate: A key health indicator for a country's economy. [Online] Available at: https://www.businessinsider.com/personal-finance/unemployment-rate?r=US&IR=T [Accessed 06 December 2022]. Lee, D., 2021. How To Calculate Unemployment Rate (And Why It's Important). [Online] Available at: https://www.indeed.com/career-advice/career-development/how-is-unemploymentrate-calculated [Accessed 06 December 2022]. Sharfman, A., 2022. Office for National Statistics. [Online] Available at: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/families/articl es/familiesandhouseholdsstatisticsexplained/- [Accessed 04 December 2022]. Statista, 2018. Statista. [Online] Available at: https://www.statista.com/statistics/-/average-number-bedrooms-new-britishhouses-/ [Accessed 06 December 2022]. Watt-Watchers, 2022. Activity: Population Math. [Online] Available at: www.watt-watchers.com [Accessed 06 December 2022]. Wikipedia, 2022. Wikipedia. [Online] Available at: https://en.wikipedia.org/wiki/Sith [Accessed 04 December 2022].