Below is my cleaned R data for the Completejourney data set

```{r} install.packages("tidyverse") install.packages('completejourney') library(tidyverse) library(completejourney) install.packages("naniar") library(naniar) library(lubridate) ```

```{r} promotions <- get_promotions() transactions <- get_transactions() ```

''' This section combines campaign and coupon data Campaign type "A" can be analyzed by demographics because not the same coupons were sent to everyone

campaign_details_BC: is information on all campaigns that are not Type "A" or NA (added campaign length to file)

coupon_details: is all coupon information (coupons & coupon_redemption tables) (added year and month for graphing)

campaign_redemptions: is combine coupon and campaign details for all coupons that were used (campaigns (B&C) ONLY)

campaign_dates: is just created for an inner join to do the calculation on total coupons used/life of campaign campaign_totals: is just created to be mutated for total coupons used/ life of campaign and is then mutated to separate the day number and force it to be a number

''' ```{r} campaign_details_BC <- campaigns %>% left_join(campaign_descriptions) %>% filter(campaign_type != "Type A" & !is.na(campaign_type)) campaign_details_BC <- mutate(campaign_details_BC, campaign_length = end_date - start_date) coupon_details <- coupons %>% left_join(coupon_redemptions) coupon_details <- mutate(coupon_details, redemption_year = year(redemption_date), redemption_month = month(redemption_date)) ```

```{r} campaign_redemptions <- campaign_details_BC %>% inner_join(coupon_details) %>% distinct() campaign_dates <- campaign_redemptions %>% distinct(campaign_id,campaign_length) ```

''' Total Coupons used by campaign (top 10 campaigns) '''

```{r} campaign_totals_used <- campaign_redemptions %>% group_by(campaign_id) %>% count(household_id) %>% arrange(desc(n)) %>% inner_join(campaign_dates, by = 'campaign_id') campaign_totals_used <- separate(campaign_totals_used,campaign_length, into = c('days', 'unit'), sep = ' ')

campaign_totals_used$days <- as.integer(campaign_totals_used$days) campaign_totals_used$daily_avg <- with(campaign_totals_used,n/days)

top_10_campaigns_qauntity <- head(campaign_totals_used %>% arrange(desc(daily_avg)), n = 10) ```

''' Need sales info so have to join with transactions Combining campaign redemptions with transactions

I added the inner_join to top 10 campaigns to look at top sales of products (quantity and sales_value)

'''

```{r} top_10_products_total_sales <- head(campaign_redemptions %>% inner_join(top_10_campaigns_qauntity) %>% inner_join((transactions)) %>% group_by(product_id)%>% summarize(total_sales = sum(sales_value)) %>% arrange(desc(total_sales)), n = 10)

top_10_products_quantity <- head(campaign_redemptions %>% inner_join(top_10_campaigns_qauntity) %>% inner_join((transactions)) %>% group_by(product_id)%>% summarize(total_quantity = sum(quantity)) %>% arrange(desc(total_quantity)), n = 10)

```

''' Finally, I am joining the top 10 products (by sales and quantity) to promotions so we can see the top display locations and mailer locations '''

```{r} top_10_promotions_step1 <- campaign_redemptions %>% inner_join(top_10_campaigns_qauntity) %>% inner_join(transactions) top_10_promotions_step1 <- mutate(top_10_promotions_step1,week = week(transaction_timestamp)) top_10_promotions_step1 %>% inner_join(promotions)

I need help answering the below 3 questions by graphing them out.

1.what campaigns are best from a coupon usage standpoint and total sales standpoint (Show one graph that answers the question)

2.what makes a campaign good (successful display locations and successful mailer location)(Show one graph that answers the question)

3. What campaigns are most successful with certain demo groupsShow one graph that answers the question)