Advanced data analysis and insights

- [Instructor] All right, now that you've used Copilot to enrich, clean, and even visualize your data, let's look at some ways to generate and summarize insights using a combination of these methods. The exercise file for this video is 0302 advanced data analysis. Let's open Copilot and let's say I'm an analyst at this organization. We've got sales data, and my task is to analyze the data for identifiable trends and relationships. I'll get started here. We do have some suggestions as usual. Some of them may have to do with data analysis, but I'm going to get right to it. And let's ask to visualize total sales by month of order date. And as we've been seeing, Copilot is really enjoying its Python outputs. Maybe you're not so much, so let's prompt right here. Don't use Python in this way. We'll get more traditional Excel tools back, like pivot charts and tables and so forth. Okay, we're getting a nice looking line chart here. I'm going to add this to a new sheet. You may or may not get Copilot to help you format and modify this things like changing the access labels and adding the chart. If you want help with that or if you want to try to get help, I would suggest adding that to your original prompt. Once these results get placed in the Excel workbook, it gets a little harder to have Copilot kind of communicate with and make those changes. But let's go back to our original data set and try another use case here. Let's say I'm interested in digging into sales for a specific month. I want to get total sales summarized by product category for just June, 2016 of the order date. So we'll add our prompt here. And again, I'm going to say don't use Python. We'll see what kind of results we can get. Okay, and we actually still got Python there. I'm going to say try again, don't use Python, and that may or may not do the trick. If we're still struggling, we can either just paste this entire prompt in again. Okay, that looks to be useful here. Okay. We are getting a somewhat complicated function here, and this might be a place where can we use pivot tables to see if this will do the trick. Okay, and this looks like one place where we may need to compromise and adjust. We could restart Copilot and try this again. Maybe you got something different, but we do have options at least. I'm going to go ahead and move on. And so far, we've been analyzing sales as a trend and by specific categories. Let's combine this and see if we can find the top selling category for each month of sales in the data set. I want to be pretty specific here about what columns I want. I'm going to write them exactly as they are found in the data set. Let's say don't use Python again. So we want to really specify how this is going to be done, to what columns and so forth. Write some of total sales, be specific about the aggregation method, and it looks like we do get a result here. Let's go ahead and insert it just to see what this looks like. Now, this is just a static table. Every so often you will get that and maybe worth spot checking a couple of these rows just to make sure that this was set up correctly. So there are a lot of possibilities in the outputs that we get, but also the prompts and the kind of questions that we provide Copilot as well. All of this really hinges on understanding your data and understanding your tools to get the best results from analyzing your data.