Hi there, thank you for the useful video, I have a question, why do we need to use blended data method instead of creating “calculated field with case statement” in the data source level? Thanks
Hi Mirisse! Great question! The main reason we need to use blended data here is because we want need to reaggregate the data. So.. to explain further, Dataset A 👉The first level of grouping is where we sum up the total sales by product sku. Dataset B 👉From the the aggregated dataset, we can then group the sku based on the total sales -> this is the second level of grouping or known as reaggregation in data studio documentation. We can use case when to achieve this provided that the imported data has already aggregated that each row represent single SKU and their total sales. If you observe, the raw dataset we using in this example, each row is representing a single transaction. So we have to group them first by SKU before we can create grouping with the total sales. Hope it makes sense 😊
@@mirissenadisha You're welcome! Thank you for your love and generous support ❤️ Feel free to throw your ideas to me so I know what you all love and making more of those! PS: Follow my instagram for more quick tips hehe! ❤️ I also share more behind-the-scene there :)
Hello, Danalyser, good afternoon, I have a huge question and I haven't found a solution anywhere, could you help me, please? I have a data source with the demand forecast and the daily result for a series of transport modes (bus, subway, train, etc). I added a filter to make visualizations of these results by week. For example, for week 1 I have 7 predictions and 7 outcomes (in my data source) for each vehicle type and I want to measure my weekly deviation into predicted and realized. If my weekly deviation was less than 50% I would like it to be counted in a scorecard, if it was between 50 and 75 in another scorecard and so on, until I have the n modals distributed in the boxes I created. The problema is that datastudio looks at the days individually, so if it has different daily classifications, it ends up being counted more than once in different scorecards. That is, if on day 1 the deviation in the bus demand forecast was 45% it goes into a box, but if on day 4 the deviation was 65% it is again accounted for in another box, when in fact I would like the datastudio added up the predictions and results of the 7 days and only then did the classification (like an aggregated value). I did something like: COUNT_DISTINCT(IF(real/expected-1 > 0, modal, NULL)) - however this way it does individual visualization of the data and it doesn't work for me, I wanted to use the total by modal, something like an OVER (PARTITION BY modal, week ) or something like this When I try to do: COUNT_DISTINCT(IF(SUM(real)/SUM(predicted)-1 > 0, modal, NULL)) - it throws me an error saying I can't mix metrics and dimensions. The alternative I found was to bring the data consolidated by week, and then when I apply the filter to a single week at a time it works fine, but when I try to visualize two or more weeks at once it ends up falling into the same situation, it it doesn't consolidate my results by modal to then do this check, you know? Can you help me? It would help me soooo much. Thank you very much in advance! Regards from Brazil!
Hello, love your channel and learning a lot from it. I vreated a pivot chart and a date range comparisson - Do you know if its possible to do a conditional formatting on the delta (% chabged) and not the actual values? Thanks!
Hi there! Thank you for your love and support, I'm glad you find my channel useful! Based on my limited understanding, it's only possible to do conditional formatting on a dimension field / metric field. With that being said, if you are calculating the delta (%) using the default setting and not a calculated field, it is not possible. However, if you managed to have a field for the delta values, it should be able to get done. Hope this gives you some ideas to work on next :)
Hey again, do you if its possible to in a table change the row field dynamically based on a parameter the user chooses? i.e. when they choose source the table will be by source, or source/medium etc. I tried and it didnt work for me. Thanks!
Hi hi! How're you doing ? :) I came across this article that guide you on how to change the dimension based on the filter chosen by users. I may consider doing the same kind of video in my channel, but for the time being do refer to this written guide - datamonkeysite.com/2020/08/21/dynamic-change-of-dimension-using-google-data-studio/
Hi there, thank you for the useful video, I have a question, why do we need to use blended data method instead of creating “calculated field with case statement” in the data source level? Thanks
Hi Mirisse! Great question!
The main reason we need to use blended data here is because we want need to reaggregate the data.
So.. to explain further,
Dataset A 👉The first level of grouping is where we sum up the total sales by product sku.
Dataset B 👉From the the aggregated dataset, we can then group the sku based on the total sales -> this is the second level of grouping or known as reaggregation in data studio documentation.
We can use case when to achieve this provided that the imported data has already aggregated that each row represent single SKU and their total sales.
If you observe, the raw dataset we using in this example, each row is representing a single transaction. So we have to group them first by SKU before we can create grouping with the total sales.
Hope it makes sense 😊
@@danalyser thank you so much! Your explanation is clear. Love this kind of content, can’t wait to see the upcoming ones ❤️
@@mirissenadisha You're welcome! Thank you for your love and generous support ❤️
Feel free to throw your ideas to me so I know what you all love and making more of those!
PS: Follow my instagram for more quick tips hehe! ❤️ I also share more behind-the-scene there :)
Thank you the tutorials, they are amazing
Glad you like them!
Heya, I love your video super helpful. I wanted to practise but I was unable to find the data set in the description, could you share it?
Hey Jessica, thanks so much for watching. Here is the link: bit.ly/danalyser-sample-data
Let me know how your practice goes, happy learning 💛✨
Hello, Danalyser, good afternoon, I have a huge question and I haven't found a solution anywhere, could you help me, please?
I have a data source with the demand forecast and the daily result for a series of transport modes (bus, subway, train, etc). I added a filter to make visualizations of these results by week. For example, for week 1 I have 7 predictions and 7 outcomes (in my data source) for each vehicle type and I want to measure my weekly deviation into predicted and realized. If my weekly deviation was less than 50% I would like it to be counted in a scorecard, if it was between 50 and 75 in another scorecard and so on, until I have the n modals distributed in the boxes I created. The problema is that datastudio looks at the days individually, so if it has different daily classifications, it ends up being counted more than once in different scorecards. That is, if on day 1 the deviation in the bus demand forecast was 45% it goes into a box, but if on day 4 the deviation was 65% it is again accounted for in another box, when in fact I would like the datastudio added up the predictions and results of the 7 days and only then did the classification (like an aggregated value).
I did something like:
COUNT_DISTINCT(IF(real/expected-1 > 0, modal, NULL)) - however this way it does individual visualization of the data and it doesn't work for me, I wanted to use the total by modal, something like an OVER (PARTITION BY modal, week ) or something like this
When I try to do:
COUNT_DISTINCT(IF(SUM(real)/SUM(predicted)-1 > 0, modal, NULL)) - it throws me an error saying I can't mix metrics and dimensions.
The alternative I found was to bring the data consolidated by week, and then when I apply the filter to a single week at a time it works fine, but when I try to visualize two or more weeks at once it ends up falling into the same situation, it it doesn't consolidate my results by modal to then do this check, you know?
Can you help me? It would help me soooo much. Thank you very much in advance! Regards from Brazil!
Hello, love your channel and learning a lot from it. I vreated a pivot chart and a date range comparisson - Do you know if its possible to do a conditional formatting on the delta (% chabged) and not the actual values? Thanks!
Hi there! Thank you for your love and support, I'm glad you find my channel useful!
Based on my limited understanding, it's only possible to do conditional formatting on a dimension field / metric field. With that being said, if you are calculating the delta (%) using the default setting and not a calculated field, it is not possible. However, if you managed to have a field for the delta values, it should be able to get done.
Hope this gives you some ideas to work on next :)
@@danalyser yes thanks a lot
@@avishaiz you’re welcome
Hey again, do you if its possible to in a table change the row field dynamically based on a parameter the user chooses? i.e. when they choose source the table will be by source, or source/medium etc. I tried and it didnt work for me. Thanks!
Hi hi! How're you doing ? :)
I came across this article that guide you on how to change the dimension based on the filter chosen by users. I may consider doing the same kind of video in my channel, but for the time being do refer to this written guide - datamonkeysite.com/2020/08/21/dynamic-change-of-dimension-using-google-data-studio/
@@danalyser Thanks I will check it out