Topics started by Anders Brakestad

25 Topics

Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

Is it possible to specify whether time series data is aggregated at "start", "midpoint", or "end"?

I am making comparisons between time series data in CDF and PI. The reason is that in our tenants the CDF data is not 100% accurate compared to PI.However, from my testing I think that PI performs its aggregations with the timestamps “centered” at the aggregated time periods, while CDF puts the timestamps at the start of each aggregated period. Is it possible to specify how this is done with the Python API? From my study of the docs it appears not to be the case. The same applies to the PI Web API as well: I cannot specify how the timestamps are placed. The agreement with PI becomes significantly better if I place the CDF timestamps at the center of the aggregated time periods.My current workaround is the following:Fetch RAW data from CDF Shift the timestamps by 0.5x of the granularity Resample to the desired granularity Compute mean Interplate any missing valuesThe issue is that fetching raw data is a lot more time consuming than fetching aggregates. I have been playing with fetching

4
Anders Brakestad
Seasoned
1 month ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Deploy streamlit app by referencing a github repoGathering Interest

Being able to deploy simple streamlit apps in CDF is very cool, but as a developer I really do not like the workflow of writing code in the CDF editor - or copy code from my development repository into the CDF editor.Similar to Cognite Functions, I want to just make sure the necessary files are there, like requirements and perhaps a standardized entry point. 

1
Lars Moastuen
Seasoned Practitioner
2 months ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

cognite-sdk for Python as a conda package

As of today I think the Python SDK is only accessible via pip. While I can install a pip package into a conda environment, I don’t really like this solution. Are you planning on releasing a conda package for the Python SDK?

2
H
Practitioner
3 months ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

Sudden error in OAuth authentication

I have a helper function that returns a CogniteClient via interactive OAuth authentication. This worked literally yesterday, but gives me an error today.This is my function definition import osfrom cognite.client import CogniteClient, ClientConfigfrom cognite.client.credentials import OAuthInteractivefrom typing import Uniondef authenticate_to_cognite_oauth( project: str="akerbp", client_name: str="Employee", client_id: Union[str, type(None)]=None, tenant_id: Union[str, type(None)]=None, base_url: str="https://api.cognitedata.com" ) -> CogniteClient: """Authenticate to Cognite via OAuth interactive login. Return cognite.client.CogniteClient.""" if client_id is None: client_id = os.getenv("CDF_CLIENT_ID") if tenant_id is None: tenant_id = os.getenv("CDF_TENANT_ID") authority_url = f"https://login.microsoftonline.com/{tenant_id}" scopes = [f"{base_url}/.default"] creds = OAuthInteractive( authority_url=authority_u

3
Anders Brakestad
Seasoned
8 months ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Dashboarding capabilities in CDFPlanned for development

I am missing a proper dashboarding functionality. A canvas where we can place different types of visuals that we connect to data form CDF. A use case relevant to us is to monitor certain equipment health over time. Various visuals are relevant: time series plots, KPI cards with “delta value” functionality (difference from previous value in card), bar charts, histograms, pie charts, status indicators (green light, orange light, red light), and so on.Is this something that has been discussed for the roadmap, or will Cognite go go into this market? There are certainly many different tools to use, such as Power BI, Grafana, or custom Python solutions with streamlit/dash, but being able to quickly set something up directly in CDF without having to deal with connectors would be very welcome.

3
C
Seasoned Practitioner
8 months ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

Obtaining access to the Github repo for InDSL

Is it possible to gain access to the Github repo for the InDSL? I will use Charts in my work, and would like to be able to contribute to the development.

6
K
Seasoned Practitioner
9 months ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: A way to explicitly set the TS granularityGathering Interest

I think the automatic granularity used in Charts works very well, and 90% of the time it selects a granularity that is appropriate. However, in some situations I would like to actually use a finer granularity than the automatic one.How exactly the UX should be for setting this, I am not sure. One option is to place a granularity selection button next to the TS color button in the table. A more fancy solution could be to bind a keyboard key to the mouse wheel, so that when ALT is held while scrolling the granularity is automatically updated and displayed in the table. Regards,Anders Brakestad

8
K
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Allow changing the y axis limits while in stacked modeGathering Interest

I often use the “stacked y axis” mode, because it tidies up the graphing area with multiple time series. However, this mode also prevents me from adjusting the y axis limits for the individual time series. This restriction needs to go, because the automatic axis limits do not always let me see the data at the detail necessary. There is no logical reason as I can see why I should not be allowed to change the y axis limits. The stacked mode should just place the center (y_max - y_min) at the location defined by the stack algorithm, and be agnostic to the actual valuels of y_max and y_min.

4
K
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Allow users to define calculation templatesGathering Interest

It would save the user time if it was possible to store a calculation setup as a template. There could be a calculation template library (public and private) that we could select from within the calculation node editor. Perhaps a particular equation that involves five TSs, 5 different Charts functions, and 8 constants, and not having to build this from scratch every time would be very useful. One could duplicate an existing Chart, but this is not addressing the root cause. I think a Calculation Library could be useful. Regards,Anders Brakestad

2
K
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

How to ensure identical DateTimeIndex in TS data from CDF and Pi (via Seeq API)

I am trying to make a comparison between time series data from CDF and from Pi (accessed with the Seeq Python API). Some questions arise:​​​​​How do I ensure that the DateTimeIndex-s that are returned are identical? Why does CDF return DateTimeIndex with dates before my start value?Examples:Pulling a raw time series from Pi via Seeq looks like this:start = datetime(2021, 1, 1)end = datetime(2021, 1, 2)spy.pull(items, start=start, end=end, grid=None, quiet=True)And the CDF analog looks like this:start = datetime(2021, 1, 1)end = datetime(2021, 1, 2)cognite.datapoints.retrieve_dataframe( external_id=external_id, start=start, end=end, granularity=None)The resulting DateTimeIndex-es are not identical:From Pi via Seeq:DatetimeIndex(['2021-01-01 00:00:00+01:00', '2021-01-01 00:00:05+01:00', '2021-01-01 00:00:10+01:00', '2021-01-01 00:00:15+01:00', '2021-01-01 00:00:20+01:00', '2021-01-01 00:00:25+01:00', '2021-01-01 00:00:30+01:00', '2021

7
Anders Brakestad
Seasoned
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

Feedback and feature requests for Asset Data Insight (and Cognite Charts)

Asset Data Insight (ADI) looks to be a very promising analysis tool. I realize it is in beta, but I want to share a couple of my impressions from using it over the last couple of days. Perhaps these suggestions can make it to a list of features to be implemented for the coming releases.I am a bit confused about the future development of ADI and Cognite Charts. There is clear overlap in functionality, and yet key features in ADI are not present in Charts, and vice versa. Will one take over for the other? Will they coexist? If one will be deprecated in the future, my feedback still is relevant for the one that survives.#1 Make all Chart functionality available in ADI.The calculation node editor is especially useful, but the interface for customizing how toe visualization looks is better in Charts than in ADI. #2 Improve the Events layerOverlaying events, such as maintenance history, on key time series can be very useful. However, the current implementation is lacking. The filtering of ev

9
I
Seasoned
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Apply formatting to all TSsGathering Interest

I have on several occasions have wanted to make the line widths for all TSs in the Chart larger. It gets tedious doing this manually for 10+ TSs, so why not have functionality for applying certain formatting options to multiple TSs? Line widths, colors, line styles, granularity (if implemented), etc. Regards,Anders Brakestad

3
Magdalena Rut
Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Storing different "Views" of a Chart for a user to jump betweenGathering Interest

I am often building Charts for specific purposes in root cause analyses, and I want to be able to tell stories and highlight important aspects in the data. Key aspects in the TS might be visible at different time scales and at different dates, which requires me to spend time navigating and find the relevant sections of the data.I would like to be able to save a specific view of the graphing area and give it a name. These views should be easily accessible within Charts so that I can simply click on it and the exact same view will be shown. With something like this implemented I can prepare a “story” or “presentation” directly in Charts. This way I don’t have to i) make multiple Charts to store different views, or ii) save PNGs and put these into a Power Point.All parameters needed to reproduce the view needs to be stored: active time series, colors, line widths, line styles, x limits, y limits…. anything needed to reproduce the view. Regards,Anders Brakestad

3
M
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Apply function to multiple TSsGathering Interest

If I have 10 TSs in one Chart, and I want to apply a smoothing function to all, I currently need to do this 10 times. There should be a possibility of replicating a calculation for multiple TSs. Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Export a calculation as Python codeGathering Interest

Seeing as how all functions available in a calculation is defined in Python, perhaps a complete calculation (all nodes) can be exported to an executable python code. I am just throwing this out there as an idea; I have not thought a lot whether this is difficult/useful. One use case could perhaps be to prototype something in Charts, and then use exported code as a starting point for building something in Python. Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Highlight a TS when hovering over its y axisPlanned for development

With 15 time series in one Chart, it can get a bit difficult to keep control over which one is which, especially when I want to modify the y axes. The colors are very similar, and I need to try a few different ones before I find the correct TS. If the TS would be highlighted in some way, e.g. bold, when I hoved over its y axis, it becomes trivial to locate the correct axis.Note that this issue is related to this one. If I could modify the axis by interacting directly with the TS, then this issue would disappear I suppose. Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Possibility of grouping together nodes in a calculation with colored boxesGathering Interest

The node based calculation editor is neat, and offers a visual way of building quite complex TS transformations and calculations. But, the “visual complexity” grows quite fast, and I am missing a basic feature common in such node editors where the user can add colored boxes with a title behind some of the nodes. In this way we get some structure and documentation into the calculation, and it becomes easier for the user (and other people as well) to quickly understand what exactly is being done.Here is an example from Blender, which uses a node editor to build complex ray tracing models: Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Keyboard bindings for more efficient useGathering Interest

As a Data Scientist I want to visualize data efficiently so that I spend less time clicking and more time focusing on the data.I really think Charts can benefit from some clever shortcuts and keyboard bindings. For example:Currently I can zoom in on the y axis by by scrolling on the relevant axis, but I find that I often hide the y axis when I load multiple TSs. It would save me time and lot of clicking if I instead could hover over the TS in the tabular view, or just on the graphed TS itself, and then SHIFT+SCROLL to change the y axis. This way I can quickly manipulate the view for many time series, while having the full graph area available for the visualizing. With many TSs the axis area takes up a lot of space I would also like to be able to shift the TS up and down by a similar keyboard binding. E.g. CTRL+DRAG to move the TS up or down in the plotting area.I am sure there are a lot of possible keyboard bindings that will make the UX better. But at the simple cases above should be

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Figure element formatting for exportGathering Interest

I have been involved in a root cause analysis lately, and we are not working on the final report. I have been asked to produce a couple of trends, and my go-to tool was Charts. While there is a PNG export functionality, I find it so limiting that I ended up in Python and matplotlib to produce a decent looking figure of a simple differential pressure time series.Some of the things I am missingPossibility to set font sizes of x and y labels Set a figure title (displayed on top of the figure) How detailed the x ticks and x tick labels should be. The user should be able to set the time interval, either with a dropdown of preset choices, or a custom string formatting option for “pro” users (“%Y-”b”) to produce labels of the form “2021-Jan”. More colors to choose from, and to set the hex manually (I especially miss “black”). To change the figure aspect ratio and dpi. To set the background color to a dark color. A legend should be autogenerated based on the currently visible time series (name

6
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Activate text box when editing the TS namePlanned for development

A textbox appears when I click the edit button for the TS name, but the textbox is not active. I need to click it to start typing. This is annoying. The only reason anyone clicks that button is to edit the name. So it should be automatically activated with the cursor at the end. Regards,Anders Brakestad

3
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: A place to share notes/documentation in a ChartPlanned for development

I think Charts need a feature for storing notes, thoughts, documentation, or other types of descriptive texts concerning the Chart. This would make it more self contained. For example, it would be useful to have a description of why certain tags have been added, what we aim to accomplish with this chart, which equation is used in the calculations, assumptions in the analysis, how to use the Chart, other related charts, etc. Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Generate new Chart of selected TSs from current ChartGathering Interest

Expectation: “I only need a few tags in this Chart, so I’ll just add everything in this one”Reality: The Chart grew to 30 TSs and I probably need to split it up into individual Charts.I would like the possibility of selecting a few TSs from the table, and populate a new Chart with the selected TSs. This should maintain the formatting, calculations, and axis views currently set. As a workaround I suppose I could duplicate the Chart and delete everything I don’t need, but this introduces unnecessary effort in cleaning up the new Chart. Regards,Anders Brakestad

2
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned

Cognite Charts: Select zoom window with the mouseImplemented

The scroll “zoom” works very well, but on several occasions, I have missed the ability to simply select a window with the mouse, which then is zoomed in to. When looking for a particular time window this method simply just works, without all the scrolling and dragging to manipulate the view to my wish.

3
Anita Hæhre
Seasoned Practitioner
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in Developer and User Community

Wildcard listing resources

I am listing time series for a given asset, and get a lot of results. I need to filter based on the time series external ID, but the only option is by using the “external_id_prefix” argument to the list function. So I build up the prefix left to right. Somewhere in the external id is a parameter that I do not care about, and after comes a new parameter that I want a particular value of. Concrete example of external IDs:IAA_Miros_Weather_Data_WIA_008IAA_Miros_Weather_Data_WIB_008IAA_Miros_Weather_Data_WIC_008IAA_Miros_Weather_Data_WID_008IAA_Miros_Weather_Data_WIE_008I am interested in only getting time series with external IDs that contain “_WI” and that end with “008”. Is there a way to list time series with a wildcard? Something like this:client.time_series.list( asset_ids=[my_asset_id], limit=None, external_id_prefix="IAA_Miros_Weather_Data_WI*_008", partitions=4)I could of course obtain the relevant time series by filtering after the fact. Something like thisresult = cl

3
Anders Brakestad
Seasoned
1 year ago
Anders Brakestad
Seasoned
Anders BrakestadSeasoned
asked in How-To Guides

How to retrieve specific Miros weather data for a given asset ID? [Community Contributed]

Hi all!I am trying to obtain weather data from Miros for a given asset ID. I am experienced in Python, but new to CDF, and don’t have access to a lot of practical examples of common tasks and patterns. So I thought I’d ask here to speed up my work :) What would be best practice for the given scenario (given in detail to try to avoid the XY problem)?I am going to work on some root cause analyses, and the starting point is likely to be an asset that failed multiple times for the same reason in the past. I want to fetch different types of data for this asset, such as maintenance history, relevant metadata, and weather data at its location during the relevant time window. This will eventually make its way into a simple graphical tool that I can use at the start of new root cause analyses.I have been playing around with retrieving an asset object based on its id or external_id. This seems to be fine. Then I notice that weather time series from Miros are connected to assets. But how would I

1
K
Practitioner
1 year ago

Badge winners

Show all badges
Terms & ConditionsCookie Settings