Fix AttributeError when design is a DataFrame (#443)

When DeseqDataSet is created with a precomputed design matrix (DataFrame)
instead of a formula string, cond() and contrast() raise a confusing
AttributeError because formulaic_contrasts is never set.

Added guard with a clear error message directing users to pass contrast
vectors directly, matching the existing pattern used for the variables
property.

Fixes #440.

Author: Ekin-Kahraman

pydeseq2/dds.py

@@ -575,11 +575,25 @@ def cond(self, **kwargs):
         ndarray
             A contrast vector that aligns to the columns of the design matrix.
         """
-        return self.formulaic_contrasts.cond(**kwargs)
+        try:
+            return self.formulaic_contrasts.cond(**kwargs)
+        except AttributeError:
+            raise AttributeError(
+                "The cond() method requires a formula-based design. "
+                "When using a precomputed design matrix (DataFrame), "
+                "pass the contrast vector directly instead."
+            ) from None
 
     def contrast(self, *args, **kwargs):
         """Get a contrast for a simple pairwise comparison."""
-        return self.formulaic_contrasts.contrast(*args, **kwargs)
+        try:
+            return self.formulaic_contrasts.contrast(*args, **kwargs)
+        except AttributeError:
+            raise AttributeError(
+                "The contrast() method requires a formula-based design. "
+                "When using a precomputed design matrix (DataFrame), "
+                "pass the contrast vector directly instead."
+            ) from None
 
     def fit_size_factors(
         self,