HyukjinKwon commented on a change in pull request #34931:
URL: https://github.com/apache/spark/pull/34931#discussion_r771915801
##########
File path: python/pyspark/pandas/frame.py
##########
@@ -8828,22 +8842,117 @@ def describe(self, percentiles: Optional[List[float]]
= None) -> "DataFrame":
else:
percentiles = [0.25, 0.5, 0.75]
- formatted_perc = ["{:.0%}".format(p) for p in sorted(percentiles)]
- stats = ["count", "mean", "stddev", "min", *formatted_perc, "max"]
+ if len(exprs_numeric) == 0:
+ if len(exprs_non_numeric) == 0:
+ raise ValueError("Cannot describe a DataFrame without columns")
- sdf = self._internal.spark_frame.select(*exprs).summary(*stats)
- sdf = sdf.replace("stddev", "std", subset=["summary"])
+ # Handling non-numeric type columns
+ # We will retrive the `count`, `unique`, `top` and `freq`.
+ sdf = self._internal.spark_frame.select(*exprs_non_numeric)
- internal = InternalFrame(
- spark_frame=sdf,
- index_spark_columns=[scol_for(sdf, "summary")],
- column_labels=column_labels,
- data_spark_columns=[
- scol_for(sdf, self._internal.spark_column_name_for(label))
- for label in column_labels
- ],
- )
- return DataFrame(internal).astype("float64")
+ # Get `count` & `unique` for each columns
+ has_timestamp_type = True in is_timestamp_types
+ if not has_timestamp_type:
+ counts, uniques = map(
+ lambda x: x[1:], sdf.summary("count",
"count_distinct").take(2)
+ )
+ else:
+ # `summary` doesn't support for timestamp column, so we should
manually compute it
+ # if timestamp type column exists.
+ counts = []
+ uniques = []
+ exprs = []
+ for column in exprs_non_numeric:
+ exprs.append(F.count(column))
+ exprs.append(F.count_distinct(column))
+
+ count_unique_values =
list(sdf.select(*exprs).first().asDict().values())
+ for i in range(0, len(count_unique_values) - 1, 2):
+ counts.append(str(count_unique_values[i]))
+ uniques.append(str(count_unique_values[i + 1]))
+
+ # Get `top` & `freq` for each columns
+ tops = []
+ freqs = []
+ for column in exprs_non_numeric:
+ top, freq = (
+ sdf.groupby(column)
+ .count()
+ .sort("count", ascending=False)
+ .first()
+ .asDict()
+ .values()
+ )
+ tops.append(str(top))
+ freqs.append(str(freq))
+
+ stats = [counts, uniques, tops, freqs]
+ stats_names = ["count", "unique", "top", "freq"]
+
+ # Get `first` & `last` for each columns if timestamp type column
exists.
+ if has_timestamp_type:
+ exprs = []
+ for is_timestamp_type, column, column_name in zip(
+ is_timestamp_types, exprs_non_numeric, column_names
+ ):
+ if is_timestamp_type:
+ # `first` & `last` are min & max respectively for
timestamp type.
+ exprs.append(F.min(column))
+ exprs.append(F.max(column))
+ else:
+ # `first` & `last` are always NaN for string type
columns.
+ exprs.append(F.lit(np.NaN).alias(column_name +
"_first"))
+ exprs.append(F.lit(np.NaN).alias(column_name +
"_last"))
+
+ firsts = []
+ lasts = []
+ first_last_values =
list(sdf.select(*exprs).first().asDict().values())
Review comment:
btw, can we avoid using `asDict().values()`? this is unnecessary. you
can just get it from `list(row)`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
