Filtering for Top Across Panes

本笔记只做复习以及巩固知识点使用,初次学习请下载练习工作簿,跟随网络课程的讲解同步操作,结果参考解法工作簿

Filtering for the Top 10

If we want to filter for the top 10 products by sales per region, this analysis can be interpreted two ways:

  1. Of our overall top 10 products, how do those 10 rank in each region?
    or
  2. In each region, which are the top 10 products?

The first is fairly simple, and is shown here.

Right clicking on the Product ID filter and selecting Edit Filter, then clicking the Top tab, we see this is Top 10 Product IDs by sum of sales.

But if we’re trying to answer the second scenario, “in each region, which are the top 10 products?”, we might be tempted to build the view we have here, then to put Region onto filter and select whatever Region we’re interested in.

However, this view appears to be giving incorrect responses – despite the fact our Product ID filter says top 10 by Sales, we only have 7 rows. This is a matter of how the filter is being evaluated.

If you’re familiar with SQL, filters are written in the WHERE of the query clause via an AND statement. Therefore, records need to meet all criteria to be returned – For the bar chart here on top, results need to be members of the top 10 of all Products by sales.

And if we filter by Region as well products ALSO need to be sold in that Region. If 3 of the top 10 products aren’t sold in Eastern Africa, the result subset will only contain 7 products. Both conditions of top 10 and region need to be met simultaneously for a product to be returned.

To filter for the top 10 products in a specific region, the setup is a little different. We need to let Tableau know that the Region needs to be considered before the Top 10 is calculated.

We could put the Region as a context filter, like we currently have for Market, but that isn’t very easily changed if we want to look at Northern Africa, then Western, etc.

Top 10 Across Dimensions Using INDEX

Alternatively, we can use the INDEX calculation to assign each product an index value per region, and then we can filter on this value.

Right click in the data pane and select Create Calculated Field. We’ll name this “Index”. In the formula area, type “INDEX()”. Click ok.

Find the new field, Index, here in the data pane. It’s currently a continuous measure. We’ll right click and select “convert to discrete”. This step is vital, because we need the index values to function as labels. Bring the Index field between Region and Product ID.

Now we need to tell Tableau how to calculate the Index—in other words, how we want the index to be assigned. Right click on the Index pill and select “Edit Table Calculation”. We’ll click Specific Dimensions, giving us complete control. We’ll leave Region and Product ID checked, keeping them as addressing fields because we want both to be considered when computing the index. We’ll set the “at the level” to Product ID, because we want the index values to be assigned per Product ID. And we’ll set “Restarting Every” to Region because we want to recalculate the index assignments independently per region. Finally, we need to set the sort. We’ll change it to Custom, and make it a Descending Sum of Sales.

Now that we have our indexes properly assigned, we can filter. We edited the table calculation on this pill specifically, so we have to bring a copy of this pill to the filter shelf. We can do so by holding down the control key while we click and drag, and we’ll select just 1-10. Now we have the top ten highest selling products individually per region.

Top 10 Across Dimensions Using Rank

Similarly, we could do the same process using the table calculation Rank. Right click in the data pane and select Create Calculated field.

We’ll use the RANK function, and have it rank on SUM(Sales), nd we can specify we want it to be descending. RANK(SUM(Sales), ‘desc’). We can see that what we’re ranking on is very similar to how we sorted our index values.

Like with INDEX, we want to convert this to discrete. And bring it between Region and Product. Previously, we said restarting every region. Here, we’ll right click and set the Compute Using to Pane down. This essentially does the same thing. We’ll again bring this pill to filter, select none, and choose 1-10. And we see this is an alternative way to again get the top 10 sales per region.

最后编辑于
©著作权归作者所有,转载或内容合作请联系作者
  • 序言:七十年代末,一起剥皮案震惊了整个滨河市,随后出现的几起案子,更是在滨河造成了极大的恐慌,老刑警刘岩,带你破解...
    沈念sama阅读 220,809评论 6 513
  • 序言:滨河连续发生了三起死亡事件,死亡现场离奇诡异,居然都是意外死亡,警方通过查阅死者的电脑和手机,发现死者居然都...
    沈念sama阅读 94,189评论 3 395
  • 文/潘晓璐 我一进店门,熙熙楼的掌柜王于贵愁眉苦脸地迎上来,“玉大人,你说我怎么就摊上这事。” “怎么了?”我有些...
    开封第一讲书人阅读 167,290评论 0 359
  • 文/不坏的土叔 我叫张陵,是天一观的道长。 经常有香客问我,道长,这世上最难降的妖魔是什么? 我笑而不...
    开封第一讲书人阅读 59,399评论 1 294
  • 正文 为了忘掉前任,我火速办了婚礼,结果婚礼上,老公的妹妹穿的比我还像新娘。我一直安慰自己,他们只是感情好,可当我...
    茶点故事阅读 68,425评论 6 397
  • 文/花漫 我一把揭开白布。 她就那样静静地躺着,像睡着了一般。 火红的嫁衣衬着肌肤如雪。 梳的纹丝不乱的头发上,一...
    开封第一讲书人阅读 52,116评论 1 308
  • 那天,我揣着相机与录音,去河边找鬼。 笑死,一个胖子当着我的面吹牛,可吹牛的内容都是我干的。 我是一名探鬼主播,决...
    沈念sama阅读 40,710评论 3 420
  • 文/苍兰香墨 我猛地睁开眼,长吁一口气:“原来是场噩梦啊……” “哼!你这毒妇竟也来了?” 一声冷哼从身侧响起,我...
    开封第一讲书人阅读 39,629评论 0 276
  • 序言:老挝万荣一对情侣失踪,失踪者是张志新(化名)和其女友刘颖,没想到半个月后,有当地人在树林里发现了一具尸体,经...
    沈念sama阅读 46,155评论 1 319
  • 正文 独居荒郊野岭守林人离奇死亡,尸身上长有42处带血的脓包…… 初始之章·张勋 以下内容为张勋视角 年9月15日...
    茶点故事阅读 38,261评论 3 339
  • 正文 我和宋清朗相恋三年,在试婚纱的时候发现自己被绿了。 大学时的朋友给我发了我未婚夫和他白月光在一起吃饭的照片。...
    茶点故事阅读 40,399评论 1 352
  • 序言:一个原本活蹦乱跳的男人离奇死亡,死状恐怖,灵堂内的尸体忽然破棺而出,到底是诈尸还是另有隐情,我是刑警宁泽,带...
    沈念sama阅读 36,068评论 5 347
  • 正文 年R本政府宣布,位于F岛的核电站,受9级特大地震影响,放射性物质发生泄漏。R本人自食恶果不足惜,却给世界环境...
    茶点故事阅读 41,758评论 3 332
  • 文/蒙蒙 一、第九天 我趴在偏房一处隐蔽的房顶上张望。 院中可真热闹,春花似锦、人声如沸。这庄子的主人今日做“春日...
    开封第一讲书人阅读 32,252评论 0 23
  • 文/苍兰香墨 我抬头看了看天上的太阳。三九已至,却和暖如春,着一层夹袄步出监牢的瞬间,已是汗流浃背。 一阵脚步声响...
    开封第一讲书人阅读 33,381评论 1 271
  • 我被黑心中介骗来泰国打工, 没想到刚下飞机就差点儿被人妖公主榨干…… 1. 我叫王不留,地道东北人。 一个月前我还...
    沈念sama阅读 48,747评论 3 375
  • 正文 我出身青楼,却偏偏与公主长得像,于是被迫代替她去往敌国和亲。 传闻我的和亲对象是个残疾皇子,可洞房花烛夜当晚...
    茶点故事阅读 45,402评论 2 358

推荐阅读更多精彩内容