Can we configure an action on a Present so that, when it navigates to a Discover via actions, the slicers are passed through as columns in the destination Discover?

Thanks,

Raveena

I want to collate multiple discovers onto a single page and export them into one Excel file. The export should appear as shown here, where each table represents a separate discover.

 I tried adding all the discovers into a presentation slide, but when I export, they end up in different sheets. How can I achieve the desired output shown above?

Thanks,

Raveena

I want to collate multiple discovers onto a single page and export them into one Excel file. The export should appear as shown here, where each table represents a separate discover. I tried adding all the discovers into a presentation slide, but when I export, they end up in different sheets. How can I achieve the desired output shown below.

Thanks,

Raveena

I am trying to export a Publication to Excel and want to maintain the layout of my page. Specifically: I have three reports/grids on a single Publication page, but when exporting to Excel, Pyramid creates three separate worksheets (tabs) instead of putting everything onto one single sheet.

What I have already tried without success:

• Grouping: Grouped all visuals on the page.

• Export Settings: Checked "Merge Common Cells" and toggled various Slicer options.

My Goal: The content of one Publication page should result in exactly one Excel worksheet, preserving the vertical arrangement of the reports.

Is there a specific setting or a "Layout-Fidelity" toggle I am missing to force Pyramid to treat the page as a single export entity?

Thanks for your help!

I found an inconsistency between the PQL documentation and the actual behavior of DateDiff in a Custom Column.

https://help.pyramidanalytics.com/2024/Content/Root/developer/reference/fx/PQL/Common/Date-Time/DateDiff.htm?Highlight=datediff

According to the documentation for DateDiff, the supported flag for days is:
- "dy"
- "dayofyear"

However, when using this in a Custom Column, the following expression fails:

DateDiff(last_login_date, now(), "dy")

Error:
Unsupported date offset type: dy

At the same time, the following expression succeeds:

DateDiff(last_login_date, now(), "dd")

The issue is that "dd" is not mentioned in the documentation, while the documented "dy" does not work in this context.

Expected behavior:
- either "dy" should work as documented,
- or the documentation should be updated to reflect the actually supported flag values for this function in Custom Columns.

This looks like either a documentation issue or an inconsistency in the function implementation for Custom Columns.

Screenshots attached.

 Thanks.

my aim is to map a slicer over two different models. One is a OLAP Cube and the other one a InMemory model.

In both models I use the exact same Slicer "Buchungskreis ID" 

 And also the Elements are the same.

 When I try to map two simple test reports in a dashboard the slicer does not interacte with both.

Can someone help me out here? I also attached a PIE file of the Dashboard 

Thank you in advance,

Jona

CASE(
    CONTAINS([product], “Checking”), “Checking”,
    CONTAINS([product], “Savings”), “Savings”,
    CONTAINS([product], “CD”), “CD”,
    CONTAINS([product], “Auto”), “Auto Loan”,
    CONTAINS([product], “Loan”), “Loan”,
    CONTAINS([product], “Fixed”), “Loan”,
    CONTAINS([product], “Business”), “Business”,
    “Other”
)

We have a dashboard with a slicer for cost centres. The sales organisation should cascade on this so that only the cost centres of the selected sales organisation are shown.

We have currently incorporated the cost centres as a drop-down slicer. However, this is not ideal, as there are several levels in the selection that are not easily visible to the user without the hierarchy being displayed. 

 
If I change the type of slicer to a multiple tree selection, the box (marked in red) where I can tick the cascade option disappears. Why is that? 

As always: thank you for your help!
Jona

1. "SALES by PRODUCT CATEGORY" as stacked column on the primary Y-axis
2. "MARGIN" as line chart on the secondary Y-axis

So, it should look like this...

 ...but the PRODUCT CATEGORY (Color drop zone) should only have an effect on SALES, not on MARGIN. So, I want to have ONE line chart for the overall margin. 

(How) Is this possible?

THNAK YOU!!!
Micha

Hi PA team,

We would like to implement the following calculation:
We have a list of products, and we want to create a distinct count of these products.

However, for this measure, we would like the results to not be affected by the product color filter when used in a dashboard KPI or in a Discovery canvas.

In our data model, Product ID and Color are in the same table, and in Power BI there is an option to use REMOVEFILTERS inside CALCULATE to ignore specific filters for a measure.

Our question is:
👉 How can we implement similar functionality in Pyramid Analytics?
We have checked the community resources but could not find an equivalent option.

Thanks in advance for your support.

We are running Pyramid in a single‑tenant deployment on our own VMs, and I am trying to generate daily stats for:

• total active Models
• total active Presents

This is to help us track adoption trends and “work in progress” over time.

From what I understand, the content_tbl_item table stores all content objects, and item_type identifies the object type. Based on documentation and older forum references, I have the following mappings:

• item_type = 4 → Models
• item_type = 8 → Presents

To calculate the active counts, I am using these queries:

Models

SQL

SELECT COUNT(*) AS etl FROM dbo.content_tbl_item WITH (NOLOCK)

WHERE item_type = 4 AND (is_deleted = 0 OR is_deleted IS NULL)

AND is_active = 1;

Presents

SQL

SELECT COUNT(*) AS present FROM dbo.content_tbl_item WITH (NOLOCK)

WHERE item_type = 8 AND (is_deleted = 0 OR is_deleted IS NULL)

AND is_active = 1;

My questions

1. Are these item_type values correct (4 = model, 8 = present) in a single‑tenant Pyramid install?
2. Is this the correct and supported way to count active content items in Pyramid?
3. Are there any other flags or joins I should include to ensure accuracy (e.g., multi-version rows, ownership, tenant filters, etc.)?

We are currently designing our own homepage in Pyramid to better tailor it to the needs of our users. The latest reports and favourites should enable them to quickly find the most important reports again. On top of that, we use the news feed to present new features, such as new dashboards or new data models. Futhermore, users can report system malfunctions to us via the ticket system.

Overall, we are satisfied with the content but less so with the layout and appearance. For example, there is no option for a free text field. We currently use discoveries and name them according to the heading. But that shouldn't really be the case.

Does anyone else have any best practice examples or experience with how to make the home page more visually appealing?

I am grateful for any tips.

Best regards, Jona

does the iOS App support Configuration in Microsoft Intune?
Customers wants to prefill Pyramid Server URL over the configuration in Intune. 

https://learn.microsoft.com/en-us/intune/intune-service/apps/app-configuration-policies-use-ios

The attribute I would like to use for the months is called ‘[KJ Month in Year]’.

My first approach was to use a parameter for all ‘[KJ Month in Year]’ members

 I then inserted this into a list using the Month to Date function. 

 In the last step, I then insert the list into an aggregation to obtain only the cumulative value.

 However, at the end I receive an error message.

Does anyone have an alternative approach in mind that works?

Thanks and best regards, Jona

Other client tools see this as a date.  How do I get Pyramid to see it as a date?  Thanks.

 Here is the table in the SSAS model:

Thanks

after the latest major Release we often use the function to print a publication in the body of an e-mail. 

In some cases we have discoveries with many rows in it. I have attempeted to use "visuell flow" in order to avoid one report from visually overlapping another but it didn't work. 

Does anyone have any tips on how I can solve this problem instead?

Thank you & best wishes 

Jona

How can I prevent the pre-query popup from appearing on all slides when a slicer is set as a pre-query slicer on one slide?

Thanks,

Raveena

I was wondering if it's possible to enable scrolling in a "Column Chart" Type discovery, the thing is that i might have a discovery with many categories and it's hard to understand what's going on.

For Example :

 Thanks in advance!

I've had this problem several times in the last few days. Can you confirm this on your end, or should I look for the problem elsewhere?

THANK YOU & best regards,
Micha

A couple of questions - is there a way to configure the tool tip so that the column name and column location are omitted?

Secondly, how can we increase the tool tip's font size? 

I've looked through the help and Pyramid itself to no avail.

Thank you,

Bruce

Currently , while sending an email from Subscription and publication, we are unable to see the email recipients . Its difficult for the team to check with the teams whether the mail is received or not.

Is there any option to enable the all email recipients list (to/cc/bcc) in the email send from Pyramid, not just single individuals  (Both Subscription and Publication).

Regards,

Aaron Y

I'm building a product launch dashboard and struggling with PQL filter context. I need to calculate several "New vs Existing Buyers" metrics that require comparing two different time periods while respecting slicer selections.

My data structure:

• panelist_id - unique buyer identifier
• supplier, brand, sub_brand - supplier hierarchy
• class, category, sub_category - product hierarchy
• period_type - 'LAUNCH' or 'COMPARISON'
• is_launch_product - TRUE/FALSE
• sales_amount - measure

The metrics I need to build:

1. New Buyers % by Supplier/Brand/Sub_brand:

• A panelist is "existing" if they bought from the supplier/brand/sub_brand in COMPARISON period (excluding launch product)
• A panelist is "new" if they did NOT buy from the supplier/brand/sub_brand in COMPARISON period
• Formula: Count of new panelists / Count of all launch buyers
• Must work dynamically with slicers on supplier OR brand OR sub_brand OR Class Or Category Or Sub Category

2. New Buyers % by Class/Category/Sub_category:

• Same logic but for product hierarchy
• A panelist is "existing" if they bought from the class/category/sub_category in COMPARISON period (excluding launch product)
• A panelist is "new" if they did NOT buy from the class/category/sub_category in COMPARISON period
• Must work dynamically with slicers on class OR category OR sub_category 

The problem:

When I use Filter(AllMembers([panelist_id]), ...), the slicer context is completely ignored inside the Filter.

What I tried:

-- New Buyers (using Except)
Count(
    Except(
        Filter(
            AllMembers([table].[panelist_id]),
            ([measures].[sales_amount], [table].[is_launch_product].[true], [table].[period_type].[LAUNCH]) > 0
        ),
        Filter(
            AllMembers([table].[panelist_id]),
            ([measures].[sales_amount], [table].[period_type].[COMPARISON]) > 0
        )
    )
)
/
Count(
    Filter(
        AllMembers([table].[panelist_id]),
        ([measures].[sales_amount], [table].[is_launch_product].[true], [table].[period_type].[LAUNCH]) > 0
    )
)

Expected result: When supplier slicer = "Tnuva", the second Filter should return panelists who bought from Tnuva in COMPARISON (24 panelists).

Actual result: The second Filter returns ALL panelists who bought anything in COMPARISON (41 panelists), ignoring the supplier slicer entirely.

Other attempts that didn't work:

• CurrentMember([supplier]) in the tuple - still ignored
• NonEmpty() - didn't respect slicer context
• Nested Sum(AllMembers([brand]), Sum(AllMembers([sub_brand]), ...)) - didn't work
• Intersect() instead of Except() - same problem

Verified in SQL:

-- This returns the correct numbers:
-- Total launch buyers: 28
-- Existing buyers (same supplier): 24
-- New buyers: 4 (14.3%)

My questions:

1. Is there a way in PQL to make a nested Filter() respect the current slicer context? Something equivalent to DAX's CALCULATE with KEEPFILTERS?
2. If not possible with a single measure, what's the recommended approach?
   • Separate measures per hierarchy level (6 measures)?
   • Pre-calculate in database?
3. Is there a function I'm missing that can "inject" the current slicer selection into a Filter?

Thanks for any help!

Is it possible on a presentation to have the logo (image) change when the appropriate company is chosen via a slicer?

Does using NLQ with the Pyramid Internal chatbot cause any data (semantic as well) to leave my network ?

Or does this just use the semantic data from the model to answer my questions ?

We do not plan to configure any LLM as of now but wanted to check if using Pyramid Internal NLQ engine would cause data to leave our network

Thanks

Mad

Symptoms & Errors:
The export fails almost immediately, and I am seeing the following errors in the system logs:

1. Code Exception (NPE):
   java.lang.NullPointerException: Cannot invoke "api.models.printing.enums.ExportTitleContext.ordinal()" because "titleContext" is null
   at bl.printingEngine.PrintingEngineImplementors.ExcelSheetFactory.getSubTitle(ExcelSheetFactory.java:212)

2. Renderer Crash:
   ... got errors All puppeteer process tabs have died

3. Timeouts:
   On some attempts, the task simply times out (even after increasing maxTimeForProcess to 600s).

Troubleshooting attempted:
I have already tried the following tuning steps with no improvement:

• Reduced printing parallelism in Task Services.

• Increased printing/conversion memory limits (Max Active Server Size).

• Increased all relevant timeouts (Printing job timeout, Query limits).

• Restarted the Task Engine.

Please verify this on your side. This appears to be a critical bug with the Excel export functionality in the new release.

Thanks.

Thanks,

Chun 

Thank you,

Bruce

Any guidance, examples, or documentation references would be greatly appreciated.

Thanks in advance for your support.

I'm trying to create a range slicer system where the slider's min/max values change dynamically based on a period type selection.

My Use Case:

• I have a Rank column (integer) in my data that represents either week rank or month rank
• I have a Period Display column that indicates whether the data is "Weekly" or "Monthly"
• Weekly data: rank ranges from 1-102
• Monthly data: rank ranges from 1-24

What I Need: I want to create an interface where:

1. User selects "Weekly" or "Monthly" (via buttons or dropdown)
2. Based on that selection, range sliders appear showing:
   • If Weekly: Start Rank (1-102) and End Rank (1-102)
   • If Monthly: Start Rank (1-24) and End Rank (1-24)
3. These slider values filter my data by the Rank column

thanks in advance!

Greetings.

When exporting a CSV file from Discover or Presentation, the column order changes to a weird (non-alphabetical) sequence. I want the columns to export in the exact same order as created in Discover.

This works fine in Excel format, but only occurs with CSV. How can I preserve the original column order in the CSV export?

------------This is the Discover and see the order

------------After Exporting this in CSV format
 

Thank you in advance.

For example below i have 'spend on shop' selected- if i change to 'share',  it will still have the conditional formatting for 'spend on shop'

 Is there anyway around this ?

Thanks, 

imbar

We currently use a formula that recognises the current month and year based on the current date and then outputs the entire year (e.g. January 2025, February 2025, ..., December 2025). In other words, all months from January to December. However, the formula (written in MDX) is still very static.

I would like to link it to our calendar dimension and thus make it filterable within dashboards. Here we use the year and the month.

The year dimension is called in German [Kalender].[Jahre].[KJ Jahr] or, as an example for 2025, [Kalender].[Jahre].[KJ Jahr].&[2025].
The month dimension is called in German [Kalender].[Perioden].[KJ Monat Im Jahr] or, for example, for January, [Kalender].[Perioden].[KJ Monat Im Jahr].&[Januar]

Can someone help me out here?

Best regards,

Jona

I'd like to report an issue with Data Type preservation that creates a blocker when using the "Multi-Select" feature in Data Flow.

The Scenario:

1. I create a Data Model (Model A) with a column explicitly defined as Date (e.g., from a Date Range source).

2. I create a second Data Model (Model B) and use Model A as the Source.

3. I use the "Add as Multi-Select" option for the tables because I want to perform a direct pass-through of multiple tables.

The Problem:
When Model B reads Model A, the Date column is automatically converted/cast to Datetime.
Because I am using the "Multi-Select" node, I cannot add a "Convert Column" node to fix this back to "Date" (as Multi-Select nodes can only be connected directly to a Target).

 Impact:
This forces us to abandon the efficient "Multi-Select" feature and add tables individually just to fix a data type that should have been preserved correctly in the first place.

Proposed Solution:

1. Ensure that the Date data type is strictly preserved when reusing In-Memory models as a source.

2. OR: Allow basic Data Type mapping/overrides inside the Source Node properties even when "Multi-Select" is used.

Thank you.

Hi Pyramid Community,

I'm looking to create a dynamic table visualization where users can control which columns appear in the report through a slicer or parameter selection on the dashboard.

What I want to achieve:

• A multi-select slicer with options like: "Sales", "Market Share", "Growth %", "Distribution", "Units"
• Users check/uncheck which metrics they want to see
• The table dynamically shows/hides columns based on their selection
• Ideally without creating multiple separate table visuals

Use case: I'm building retail analytics dashboards where different stakeholders need different metric combinations. Rather than creating 10+ separate views, I want users to customize their view on the fly.

Questions:

1. Is dynamic column visibility based on user selection possible in Pyramid (Presentation or Discover)?
2. If yes, what's the recommended approach - Presentation Parameters, Formulate logic, conditional visibility, or something else?
3. Are there any limitations I should be aware of (performance, export behavior, etc.)?

I'm working primarily in Pyramid Presentations with Formulate-based data models.

Any guidance or examples would be greatly appreciated!

Thanks, Oren

 Is there some way to do this for a Tabular Grid?

I'd like to report a performance and rendering issue with slide previews in the Presentation viewer.

The Issue:
When opening a Presentation, the slide previews (thumbnails) often take an excessive amount of time to load (40+ seconds).

Observed Symptoms:

1. Grey Placeholders: Users see grey boxes instead of slide previews for a long time.

2. Stale/Incorrect Images: Sometimes, the viewer initially displays thumbnails from a previously opened presentation before eventually updating to the correct ones. This is very confusing for users.

Impact:
This significantly degrades the user experience. Users are left waiting or are misled by incorrect thumbnails, making navigation difficult.

Expected Behavior:
Slide previews should load reasonably quickly. If loading takes time, a proper loading indicator should be shown, and the system should never display cached thumbnails from a different presentation.

Screenshots:

• Screenshot 1: Grey placeholders (loading state). 

  

• Screenshot 2: Correct previews (after ~40 seconds).

  

  Please investigate the performance of the thumbnail generation/caching mechanism.

Thanks.

I can't find a slicer type for a numeric attribute, not sure what I'm missing.

I'd like to produce

I’m running into an issue with formatting when exporting to PDF and wanted to check whether this is expected behavior or if I’m missing a setting.

I’ve built a presentation in Pyramid where I’ve carefully styled and sized my tables and selected specific fonts. When I run the presentation, everything looks exactly as intended. However, when I print/export the same presentation to PDF, the output looks quite different and the table sizes change and the font appears to be different.

Is this considered normal behavior for PDF export in Pyramid, or is there something I need to configure to ensure the PDF matches what I see when running the presentation (for example, font embedding, page sizing, or export settings)?

Any guidance or best practices would be much appreciated.

Thanks in advance!

I have a request regarding security in Python/R scripts within Data Flows.

Currently, I haven't found a secure way to pass sensitive credentials (like DB passwords or API keys) into a script.

If I use standard variables or even OS environment variables, anyone with "Edit" permissions to the model can simply modify the script to print the variable value to the output logs, exposing the secret.

Proposed Solution:
It would be great to have a "Secure Variable" type in the Model. These variables should be:

1. Masked in the UI.

2. Securely injected into the script runtime.

3. Ideally, masked in the output logs if a user tries to print them.

If there is already a recommended way to handle secrets safely without exposing them to other model editors, please let me know.

Thanks!

We are trying to make it so that a user can select a single date and it will filter the data similar to this where clause, replacing the ? with the date:

where trans_date<=? and effective_date<=? and expiration_date>=?

I tried using a global parameter but then couldn't figure out how to get it to tie to the model values.  Any thoughts on how to do this?

I am running Pyramid version 2024.03.002.

I am facing a recurring but inconsistent stability issue where the UI completely crashes to a "Grey Screen" while I am working.

The Scenario:
It happens most often when I am managing permissions in the Content Explorer:

1. I select a top-level folder that contains a large amount of content (hundreds of items).

2. I go to the Roles tab.

3. I apply changes and click Propagate.

4. If I repeat this action or work extensively with these large folders, the interface eventually freezes and turns grey.

 My Question:

1. Has anyone else experienced these UI crashes when using "Propagate" on large folders?

2. Since this looks like a front-end crash, what specific logs should I check or capture to understand the root cause? (Browser Console logs, or specific server logs?)

Thanks!

Current formula:

Difference from Default =
IF(
    IsNAN([measures].[Events_DefaultAmount]),
    NULL,
    [measures].+[fc2a0693-8321-4537-95c4-68e0848127f4]-[measures].[Events_DefaultAmount]
)

Desired result:

ACTUAL PYRAMID OUTPUT

EXPECTED OUTPUT 

Is there a way to connect pyramid to Elasticsearch?

SourceType dimension

·         SourceType 1

·         SourceType 2

·         SourceType 3

·         SourceType 4

Event fact

·         Count of events

I would like to

1)      create a formula which calculates the count of events for all SourceTypes except for SourceType 3. 

2)      create a pivot table which uses the above created formula to display the count of events by SourceType, excluding SourceType 3

3)      use the formula with any other connected dimension

4)   if any other SourceTypes are added in the future, they would be included (in other words, need to use Except)

I created a list in Formulate as

{Except(AllMembers([SourceTypesTable].[SourceTypes]),[SourceTypesTable].[SourceTypes].[SourceType 3])}

And then used this list in another formula

Aggregate({[SourceTypeTable].[SourceType].*[272cf8c2-2350-4c51-bcd0-730258cbef37]},([measures].[EventsTable Events]))

This achieved #1 and #3 but not #2.

Thanks for the help.

I'm reporting a bug in the Calendar Slicer logic when using "Weeks" granularity.

Scenario:

1. I have a data model where data is filtered strictly "before today" (date < today()).

2. Assume "Today" is Monday. This means the dataset contains complete data up to and including Sunday (the full previous week).

3. In Discover, I add a Calendar Slicer and set the Week Start to Monday.

4. I switch the slicer granularity to "Weeks".

Actual Behavior:
The previous week (which is fully present in the data from Mon to Sun) is grayed out/disabled and cannot be selected.
It seems the slicer requires data for the current day (Monday) to exist in order to "close" and enable the previous week block.

 Expected Behavior:
Since all 7 days of the previous week are present in the data, that week should be available for selection. The presence or absence of data for the current week/day should not affect the selectability of the previous full week.

Workaround:
Changing the filter to date <= today() (including the current empty/partial day) enables the previous week in the slicer. However, this proves that the logic is flawed.

Thanks!

However I want to add two extra (in yellow) grand total columns that do not repeat for each date.

First column added as a new Grand Total is the total budget value for current month,  the second column is the previous years MTD value. Is this possible to add extra grand total column without them repeating the values under each date.

Example: If no sales occurred at 2:00 pm yesterday, the alert should not trigger. Currently, it triggers regardless of sales, but if I test with 3:00 pm when sales did happen, the alert works as expected.