Real-data sales analytics for Win_O_Win DA II Capstone (Lagos Business School, 2026). EDA, Visualisation, Hypothesis Testing, Correlation & Regression using R and Python in Quarto.

This translates HawkEars confidence scores into probabilities using species-specific logistic regression models.

DA II Case Study — Exploratory & Inferential Analytics using real sales data from Wins_O_Win Nigeria ltd Submitted to Lagos Business School, Prof Bongo Adi, April 2026.

TX_IP_21to24 %>% filter(`Admission Type` == "Elective" & has_ADRD == 1) %>% dplyr::select(RECORD_ID, has_ADRD, `Admission Type`, PRINC_DIAG_CODE) %>% melt(., id.vars = c("RECORD_ID", "Admission Type", "has_ADRD"), variable.name = "COLNAME", value.name = "DIAG_CODE") %>% left_join(., ICD_DX_TO_CCSR, by=c("DIAG_CODE"="ICD10")) %>% dplyr::select(`Admission Type`, CCSR_desc, has_ADRD) %>% group_by(`Admission Type`, has_ADRD) %>% mutate(PRINC_DIAG_CODE = fct_infreq(CCSR_desc) %>% fct_lump_n(n = 3, other_level = "Other") ) %>% dplyr::select(PRINC_DIAG_CODE, has_ADRD, `Admission Type`) %>% dplyr::rename(., "Principal Diagnosis"=1) %>% tbl_summary( missing_text = "NA", by="Admission Type", include = "Principal Diagnosis") ->elective_top3princ TX_IP_21to24 %>% filter(`Admission Type` == "Emergency" & has_ADRD == 1) %>% dplyr::select(RECORD_ID, has_ADRD, `Admission Type`, PRINC_DIAG_CODE) %>% melt(., id.vars = c("RECORD_ID", "Admission Type", "has_ADRD"), variable.name = "COLNAME", value.name = "DIAG_CODE") %>% left_join(., ICD_DX_TO_CCSR, by=c("DIAG_CODE"="ICD10")) %>% dplyr::select(`Admission Type`, CCSR_desc, has_ADRD) %>% group_by(`Admission Type`, has_ADRD) %>% mutate(PRINC_DIAG_CODE = fct_infreq(CCSR_desc) %>% fct_lump_n(n = 3, other_level = "Other") ) %>% dplyr::select(PRINC_DIAG_CODE, has_ADRD, `Admission Type`) %>% dplyr::rename(., "Principal Diagnosis"=1) %>% tbl_summary( missing_text = "NA", by="Admission Type", include = "Principal Diagnosis") ->emergency_top3princ TX_IP_21to24 %>% filter(`Admission Type` == "Trauma" & has_ADRD == 1) %>% dplyr::select(RECORD_ID, has_ADRD, `Admission Type`, PRINC_DIAG_CODE) %>% melt(., id.vars = c("RECORD_ID", "Admission Type", "has_ADRD"), variable.name = "COLNAME", value.name = "DIAG_CODE") %>% left_join(., ICD_DX_TO_CCSR, by=c("DIAG_CODE"="ICD10")) %>% dplyr::select(`Admission Type`, CCSR_desc, has_ADRD) %>% group_by(`Admission Type`, has_ADRD) %>% mutate(PRINC_DIAG_CODE = fct_infreq(CCSR_desc) %>% fct_lump_n(n = 3, other_level = "Other") ) %>% dplyr::select(PRINC_DIAG_CODE, has_ADRD, `Admission Type`) %>% dplyr::rename(., "Principal Diagnosis"=1) %>% tbl_summary( missing_text = "NA", by="Admission Type", include = "Principal Diagnosis") ->trauma_top3princ TX_IP_21to24 %>% filter(`Admission Type` == "Urgent" & has_ADRD == 1) %>% dplyr::select(RECORD_ID, has_ADRD, `Admission Type`, PRINC_DIAG_CODE) %>% melt(., id.vars = c("RECORD_ID", "Admission Type", "has_ADRD"), variable.name = "COLNAME", value.name = "DIAG_CODE") %>% left_join(., ICD_DX_TO_CCSR, by=c("DIAG_CODE"="ICD10")) %>% dplyr::select(`Admission Type`, CCSR_desc, has_ADRD) %>% group_by(`Admission Type`, has_ADRD) %>% mutate(PRINC_DIAG_CODE = fct_infreq(CCSR_desc) %>% fct_lump_n(n = 3, other_level = "Other") ) %>% dplyr::select(PRINC_DIAG_CODE, has_ADRD, `Admission Type`) %>% dplyr::rename(., "Principal Diagnosis"=1) %>% tbl_summary( missing_text = "NA", by="Admission Type", include = "Principal Diagnosis") ->urgent_top3princ TX_IP_21to24 %>% filter(`Admission Type` == "Not Available" & has_ADRD == 1) %>% dplyr::select(RECORD_ID, has_ADRD, `Admission Type`, PRINC_DIAG_CODE) %>% melt(., id.vars = c("RECORD_ID", "Admission Type", "has_ADRD"), variable.name = "COLNAME", value.name = "DIAG_CODE") %>% left_join(., ICD_DX_TO_CCSR, by=c("DIAG_CODE"="ICD10")) %>% dplyr::select(`Admission Type`, CCSR_desc, has_ADRD) %>% group_by(`Admission Type`, has_ADRD) %>% mutate(PRINC_DIAG_CODE = fct_infreq(CCSR_desc) %>% fct_lump_n(n = 3, other_level = "Other") ) %>% dplyr::select(PRINC_DIAG_CODE, has_ADRD, `Admission Type`) %>% dplyr::rename(., "Principal Diagnosis"=1) %>% tbl_summary( missing_text = "NA", by="Admission Type", include = "Principal Diagnosis") ->NA_top3princ tbl_stack(list(emergency_top3princ, elective_top3princ, urgent_top3princ, trauma_top3princ, NA_top3princ), group_header = c("Emergency", "Elective", "Urgent", "Trauma", "Not Available")) %>% modify_header(update = list( stat_1 ~ "**Frequency**")) %>% modify_caption("**Table 3. Top 3 CCSR Groupings for Principal Diagnoses of Inpatients 65+ in the RGV**") %>% as_gt() %>% gt::tab_style( style = gt::cell_text(weight = "bold"), locations = gt::cells_row_groups(groups = everything()) )