Use these default values:

co2 = 0 (kg/sack)
cost = 0 (kr/sack, $/sack, kg/sack?)
sack_size = 25 (kg)
env_impact = green

Color background of selected area
duplicate of #11

How to recreate:

1. Have some combinations created.
2. Refresh page
3. Watch plot jump around

Kan dere endre tekst på input til optimizer fra «Max pore size» til «Max pore size/crack opening» og «Average pore size» til «Average pore size/crack opening»

See app.py

 elif requst_ == BRIDGE_REQUEST:
        return bridgeRequestHandler(request.json.get("option"), request.json.get("value"))

UI from Blend

Test both optimizers

Text on the button linking to "Product summary" should be replaced with "LCM Library SharePoint"

When running the optimization, products that could not be found in the database should be displayed for the user in the same way as with manual combinations

Now, URLs are hard coded in different files

User controlled
Are there products which will never be used?
What should the default be? 5mu - 100mu?

Github Actions probably

Better to have a curve to the "left" of the optimal curve than to the "right"

See app.py.

    if requst_ == BLEND_REQUEST:
        return blendRequestHandler(request.json.get("products"))

Only green and yellow?
Should be able to filter out based on colour.

With this we could run optimizer indefinitely, and have user stop it when he's satisfied.

Also test other modes in report

Perhaps with a small warning that the input has changed since the bridge was drawn

+/- 10% of goal mass

See app.py

    elif requst_ == PRODUCT_ID_REQUEST:
        return productRequestHandler(request.json.get("id"), request.json.get("metadata"))

Will likely have to do some UI changes for the webapp to be usable embedded

BUT, should only keep solutions that are within 10% of the calculated total mass

Store in azure table?
Should be "closed" per logged in user.
No "secrets" to consider beyond basic login.

Something like;

• A few different input sets
• Run the optimizer many times on each set
• Test that the deviation is within acceptable margins ( there should never be an unusable result)

Should also test against a "known good solution"

Should be a field per product. Defaults to zero.
Tests should consider this.

Failes to export PDF for a solution

Looked into skipping Logic App.
Issues:

• Can only get access "on-behalf-of" user, which means it can't run on a schedule.
• Or give app permission to every sharePoint site
  Graph permission to single sites are being worked on

Fork and create PR https://github.com/equinor/architecturecontract/blob/master/contracts/LCMLibrary.md

See app.py

    elif requst_ == SIZE_STEPS_REQUEST:
        return sizeStepsRequestHandler()

Feature to download a report of the solution.
Must include bridge parameters and optimization result.

Options:

• PDF with tabels and graphs
• ExcelSheet? Probably not
• Just a text file - We should try and include some graphics

Automatic seems a bit risky. Suggest to user if a products ratio in the mix is too low?

Link to Sharepoint?
Some added info in Azure table per product?

See app.py

    elif requst_ == OPTIMIZER_REQUEST:
        value = request.json.get("value")
        name = request.json.get("name")
        products = request.json.get("products")
        mass = request.json.get("mass")
        weights = request.json.get("weights")
        environmental = request.json.get("environmental")
        option = request.json.get("option")
        iterations = request.json.get("max_iterations")
        size_step = request.json.get("size_steps_filter")

        return optimizerRequestHandler(value, name, products, mass, weights, environmental, option, iterations,
                                       size_step)