tSNE widget raises an error on data with only one variable

[PrimozGodec]

Describe the bug

tSNE widget raises an error on data with only one variable in X or when only one variable has non-nan data.

To Reproduce
Steps to reproduce the behavior:

1. In File widget select iris
2. Connect it to Select columns where you leave only one variable in features.
3. Connect to tSNE and it raises an error.

Orange version:
3.24

Additional context

Traceback (most recent call last):
  File "/Users/primoz/Documents/orange3/Orange/widgets/utils/concurrent.py", line 596, in _on_task_done
    super()._on_task_done(future)
  File "/Users/primoz/Documents/orange3/Orange/widgets/utils/concurrent.py", line 552, in _on_task_done
    self.on_exception(ex)
  File "/Users/primoz/Documents/orange3/Orange/widgets/utils/concurrent.py", line 492, in on_exception
    raise ex
  File "/Users/primoz/miniconda3/envs/orange/lib/python3.7/concurrent/futures/thread.py", line 57, in run
    result = self.fn(*self.args, **self.kwargs)
  File "/Users/primoz/Documents/orange3/Orange/widgets/unsupervised/owtsne.py", line 217, in run
    job(progress_callback=_progress_callback)
  File "/Users/primoz/Documents/orange3/Orange/widgets/unsupervised/owtsne.py", line 107, in compute_initialization
    task.initialization = task.tsne.compute_initialization(task.pca_projection.X)
  File "/Users/primoz/Documents/orange3/Orange/projection/manifold.py", line 454, in compute_initialization
    X, self.n_components, random_state=self.random_state
  File "/Users/primoz/miniconda3/envs/orange/lib/python3.7/site-packages/openTSNE/initialization.py", line 58, in pca
    embedding = pca_.fit_transform(X)
  File "/Users/primoz/miniconda3/envs/orange/lib/python3.7/site-packages/sklearn/decomposition/_pca.py", line 369, in fit_transform
    U, S, V = self._fit(X)
  File "/Users/primoz/miniconda3/envs/orange/lib/python3.7/site-packages/sklearn/decomposition/_pca.py", line 416, in _fit
    return self._fit_full(X, n_components)
  File "/Users/primoz/miniconda3/envs/orange/lib/python3.7/site-packages/sklearn/decomposition/_pca.py", line 435, in _fit_full
    % (n_components, min(n_samples, n_features)))
ValueError: n_components=2 must be between 0 and min(n_samples, n_features)=1 with svd_solver='full'