Graphing trending

[neskiz]

Hey, hoping to get some help. I have data that is input from a team weekly, and I am looking to visualize that data in a trend over time. I have not been able to get a good graphic of the data, and hoping to get some additional help. I am looking for something like a clustered column chart, but not sure that would be the best graphic since I have several rows in my data. Any insights would help a lot. thank you 

Metrix	March, 7 2025	march 14 2025
Total source 1	20625	20411
Total critical	4787	4778
active	107	123
lost	779	823
unhealthy	115	114
poor communication	14	9
bad setup	21	9
less than 50	45	46
Critial Wth issue	0	0
Critical not report	0	0
Critical bad cat 2	0	0
critical bad cat 3	0	0
Critical bad 4	16	19
critical bad 5	2	3
	1132	1180
	18	22
Overall Health percentage	97.4	97.28
Overall up down from average	0.2854	0.1627

[v-tejrama · Accepted solution]

Hi @neskiz,

Thanks for reaching out to the Microsoft fabric community forum.

I successfully created a dynamic trend visualization in Power BI using weekly input data. Initially, formatting the data for trend analysis was a challenge, so I want to outline the exact process I used for anyone facing a similar issue.

The original dataset had metrics such as Total source, Total critical, Active, Lost, and Bad setup in rows, with dates like March 7, 2025, and March 14, 2025 as column headers. To make this data suitable for trend charts, I imported it into Power BI and used Power Query to unpivot the date columns. This transformed the data from a wide to a long format, placing each date into a new Date column and corresponding values into a Metric Value column, while keeping metric names in a single Metric column.

This setup makes tracking changes over time straightforward, even across multiple KPIs.

With this structure, I built a line chart the X-axis is Date, the Y-axis is metric values, and the Metric column serves as a slicer to compare or isolate metrics. The report is now highly scalable, automatically updating as new weekly data is added.

This format also works well with clustered column or combo charts for visualizing multiple metrics. I’ve validated it with metrics like Overall Health %, Lost, and Active, and the week-over-week trends are clear.

This approach delivered excellent results, and the Power Query transformation was essential. I’m confident this method will help others streamline their reporting process.

Please find the attached pbix file for your reference.

If the response has addressed your query, please Accept it as a solution and give a 'Kudos' so other members can easily find it.

Best Regards,
Tejaswi.
Community Support

[v-tejrama]

Hi @neskiz ,

I wanted to check if you had the opportunity to review the information provided. Please feel free to contact us if you have any further questions.

Thank you.

[neskiz]

Thank you for taking the time. I have not been able to replicate this yet, but I am close. Just have a few issues I am working with on the data cleanup. Thank you for your detailed explanation 

[rohit1991]

Hi @neskiz

Could you please follow below steps :

1. Transform Data >> select the “Metric” column >> Unpivot Other Columns.

2. Rename: Attribute >> Week, Value >> Amount. Set types: Week = Date, Amount = Number.

3. (Optional) Add WeekStart = Date.StartOfWeek([Week], Day.Monday) for clean weekly buckets.

4. Build visuals:
   • Line chart >> X = Week (or WeekStart), Y = SUM(Amount), Legend = Metric.
   • If busy, use Small multiples with Metric, or add a Metric slicer.

5. Split scales: put counts in one chart and percentages (Overall Health %, Up/Down) in another with Y-axis 0–100.

6. If some metrics are always zero, hide them via a filter: Amount is not blank and not always 0.

Why not clustered columns? With many weeks and metrics, columns get cramped; lines (or small multiples) make the trend clear.

[rajendraongole1]

Hi @neskiz  - you have to perform some data transforms at power query to bring the data in below format. 

and you can derive the visuals like line chart to show the trend or you can try with cluster column chart as well. 

Hope this helps.

attached pbix file FYR