How to set Keyframes in Python for generated Objects with data based transformations

Question

after various failed attempts to make keyframes work to my liking in my first blender-script work I request your help 🥲

So I am working on a Dataviz:

1. I load Data into Blender and create an object.
2. Then I transform the object according to the data.
3. Now I have 4 different years, Year1 - Year4. I want to show each transformation, for each Year with keyframes (one keyframe shows one year location, scale transformation)
4. and all of the keyframes of the 150 country-Objects should start and end at the same time.

...But the mechanics of the for-loop are laughing at my face - I got everything working except for the keyframes which drives me mad 🤪 so If I use an offset the transformations for each object are working perfectly, but they don't start at the same time - if I try and use iteration, I have neat keyframes, but no animation (I will only show Year4). How do I get out of this Mess? 😓

See the code for the Iteration-version on github: https://github.com/SaraHeitz/blenderpython

Version with offset: https://www.youtube.com/watch?v=QTyYj4ZKNBA

Version with iteration:

So in short I would love for this Dataviz to animate the country for each year, and set a keyframe for the according values, like the youtube animation, but start and end at the same time for every object: https://www.youtube.com/watch?v=QTyYj4ZKNBA

Answer 1

You're using the Blender API okay, your problem is with Python logic.

Let's first group the rows of your .csv into a dict by year, so later we can lookup all rows in a given year.

import csv
import bpy


#---SCENE SETTINGS
scn = bpy.context.scene
scn.frame_start = 1


#---DATASET
with open ('test.csv') as f:
    rows = list(csv.reader(f))


#---GROUP BY YEAR
rows_by_year = {}
for row in rows:
    year = int(row[1])
    rows_by_year.setdefault(year, []).append(row)

Now let's sort all the years we just found. (This isn't necessary for the file you gave but it might be if the rows are in a different order.)

#---SORT YEARS
years = list(rows_by_year.keys())
years.sort()

Now we can iterate through and create everything

#---CREATE OBJECTS
object_map = {}
for iteration, year in enumerate(years):
    ita = iteration * 10
    frame = scn.frame_start + ita
    scn.frame_end = frame
    
    for row in rows_by_year[year]:
        name = row[0]
        x = float(row[2])
        y = float(row[3])
        z = float(row[4])

        if name not in object_map:
            bpy.ops.mesh.primitive_cylinder_add(
                enter_editmode=False,
                align='WORLD',
                location=(1, 1, 1),
                scale=(1, 1, 1),
            )
            object_map[name] = bpy.context.object
            object_map[name].name = name

        cylinder = object_map[name]
        cylinder.location = (x, y, z)
        cylinder.scale = (0.5, 0.5, z)
        cylinder.keyframe_insert(data_path="location", frame=frame)
        cylinder.keyframe_insert(data_path="scale", frame=frame)

We iterate first over the years, then process all rows for that year using rows_by_year. The frame is calculated just once for each year, then all rows for that year will use that frame. Then for each row in that year, we create the cyclinder (if it doesn't exist yet), and set its value and keyframe the same way you were doing.