[SPARK-49387][PYTHON] Fix type hint for accuracy in percentile_approx and approx_percentile

[zhengruifeng]

What changes were proposed in this pull request?

Fix type hint for accuracy in percentile_approx and approx_percentile

Why are the changes needed?

float accuracy is not supported:

In [9]: df.select(approx_percentile("value", [0.25, 0.5, 0.75], 1.1).alias("quantiles")).show()

...


AnalysisException: [DATATYPE_MISMATCH.UNEXPECTED_INPUT_TYPE] Cannot resolve "approx_percentile(value, array(0.25, 0.5, 0.75), 1.1)" due to data type mismatch: The third parameter requires the "INTEGRAL" type, however "1.1" has the type "DOUBLE". SQLSTATE: 42K09;

Does this PR introduce any user-facing change?

yes, minor doc change

How was this patch tested?

CI

Was this patch authored or co-authored using generative AI tooling?

No

[HyukjinKwon]

Merged to master.