I have data in the form of tables, which I wish to plot as 3D spheres - where the point would represent a sphere of radius r = value of the point.
Eg: point=4.7 --> sphere of radius 4.7
Example:
Table 1, would result in 15 spheres, at say height z=1:
| 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|
| 4.5 | 4.9 | 4.9 | 4.9 | 4.7 |
| 4.5 | 4.8 | 4.8 | 4.8 | 4.8 |
| 4.3 | 4.7 | 4.7 | 4.9 | 4.9 |
Table 2, would result in another 15 spheres shifted "upward", at say height z=2:
| 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|
| 4.3 | 4.7 | 4.7 | 4.9 | 4.7 |
| 4.4 | 4.8 | 4.8 | 4.8 | 4.7 |
| 4.5 | 4.9 | 4.9 | 4.9 | 4.8 |
I was eventually able to figure this out, the key idea was to use the marker parameter, which in-turn has a size property (ref: plotly documentation)
Example Code:
Convert the dataframes into list of numpy arrays, using the Pandas to_numpy() method and then flattening each array using the NumPy flatten() method.
For example: arrList.append(df.iloc[0:5,:].to_numpy().flatten()).
NOTE: My data also contains blanks in the form of "-".
arrList =
[array([2.5, 2.7, 3.9, 3.8, 3.9, 2.6, 2.5, 2.5, 3.9, 3.7, 2.4, 2.6, 2.4, 4,
3.9, 2.5, 2.3, 2.3, 3.9, 3.7, 3.8, 3.9, 3.6, 3.7, 3.8, 3.7, 3.7,
3.7, 3.8, 3.9], dtype=object),
array([3.8, 3.9, 2.7, 3, 2.6, 3.9, 3.3, 2.9, 2.7, 3.8, 4, 3.6, 3.9, 3.8,
3.9, 3.7, 3.8, 4, 3.9, 3.6, 3.8, 3.9, '-', '-', '-', 3.9, 3.9, '-',
'-', '-'], dtype=object)]
Since the size property assigns a size in pixels, I have defined a method to "increase" the pixels by x-times (3x in my case), for float and/or int values in my data
def sizeMask(s):
sphereSize = lambda s: s if isinstance(s, float) else s if isinstance(s, int) else 0.0
vec_sphereSize = np.vectorize(sphereSize)
return vec_sphereSize(s)
sizeList = [sizeMask(arr)*3.0 for arr in arrList]
sizeList =
[array([ 7.5, 8.1, 11.7, 11.4, 11.7, 7.8, 7.5, 7.5, 11.7, 11.1, 7.2,
7.8, 7.2, 12. , 11.7, 7.5, 6.9, 6.9, 11.7, 11.1, 11.4, 11.7,
10.8, 11.1, 11.4, 11.1, 11.1, 11.1, 11.4, 11.7]),
array([11.4, 11.7, 8.1, 9. , 7.8, 11.7, 9.9, 8.7, 8.1, 11.4, 12. ,
10.8, 11.7, 11.4, 11.7, 11.1, 11.4, 12. , 11.7, 10.8, 11.4, 11.7,
0. , 0. , 0. , 11.7, 11.7, 0. , 0. , 0. ])]
Finally, the plot is generated as follows:
x = np.asarray([[i]*5 for i in range(1,7)]).flatten()
y = np.asarray([np.arange(1,6)]*6).flatten()
data=[]
for i,arr in enumerate(arrList):
data.append(go.Scatter3d(x=x, y=y,
z=np.asarray([i+1]*30),
mode='markers',
marker=dict(size=sizeList[i], showscale=False)))
fig = go.Figure(data=data)
fig.show()
