[BUG] ModelErrorAnalysis display is empty when there is no meaningful splits

[TheSolY]

The expected behavior is to display a message such as "No meaningful properties found" Current display:

Code to reproduce:

import numpy as np
from deepchecks.vision.datasets.classification import mnist
from deepchecks.vision.checks import ModelErrorAnalysis

model = mnist.load_model()
train_ds = mnist.load_dataset(train=True, object_type='VisionData')
test_ds = mnist.load_dataset(train=False, object_type='VisionData')

def med_prop(batch):
    return [np.median(x) for x in batch]

def min_prop(batch):
    return [np.min(x) for x in batch]

new_properties = [{'name': 'median brightness', 'method': med_prop, 'output_type': 'numerical'},
                  {'name': 'min brightness', 'method': min_prop, 'output_type': 'numerical'}]

train_ds.image_properties = new_properties
test_ds.image_properties = new_properties

check = ModelErrorAnalysis(image_properties=new_properties, min_error_model_score=-0.1)
result = check.run(train_ds, test_ds, model)
result.save_as_html()

[noamzbr]

Should wait for implementation of #1780