RAND Error: ValueError: high <= 0

[GerbilPL]

Generation of new data is hindered by these errors, with nothing changed and requirements met.
Traceback (most recent call last):
File "gen.py", line 156, in
print(gen.generate())
File "gen.py", line 147, in generate
**self._generate_endpoints(),
File "gen.py", line 41, in _generate_endpoints
endpoints[endpoint] = np.random.randint(self.example_data[endpoint].max(), size=self.rows)
File "numpy/random/mtrand.pyx", line 782, in numpy.random.mtrand.RandomState.randint
File "numpy/random/_bounded_integers.pyx", line 1334, in numpy.random._bounded_integers._rand_int64
ValueError: high <= 0
This error is true for each rand call.
With rand error corrected:
It also should be noted that when some columns are empty, this can happen:
Traceback (most recent call last):
File "gen.py", line 156, in
print(gen.generate())
File "gen.py", line 147, in generate
**self._generate_endpoints(),
File "gen.py", line 57, in _generate_endpoints
endpoints[endpoint] = np.random.randint(0,self.example_data[endpoint].max()+1, size=self.rows)
File "numpy/random/mtrand.pyx", line 782, in numpy.random.mtrand.RandomState.randint
File "numpy/random/_bounded_integers.pyx", line 1324, in numpy.random._bounded_integers._rand_int64
ValueError: cannot convert float NaN to integer
Adding a zero to the problematic columns, fixes this ( so almost all ,, null ones )
history_fixed.csv
history_nan_err.csv
Proposed quick fix will be in pull requests