DEPR: deprecate DataFrame.from_csv (GH4191)

Author: jorisvandenbossche

doc/source/api.rst

@@ -938,7 +938,6 @@ Serialization / IO / Conversion
 .. autosummary::
    :toctree: generated/
 
-   DataFrame.from_csv
    DataFrame.from_dict
    DataFrame.from_items
    DataFrame.from_records

doc/source/whatsnew/v0.17.0.txt

@@ -43,6 +43,16 @@ Other API Changes
 Deprecations
 ^^^^^^^^^^^^
 
+- :meth:`DataFrame.from_csv` is deprecated. Please use :func:`pandas.read_csv`
+  instead. This function only differs from the method in some defaults:
+
+  - `index_col` is ``None`` instead of ``0`` (first column)
+  - `parse_dates` is ``False`` instead of ``True`` (parsing the index)
+
+  So a ``pd.DataFrame.from_csv(path)`` can be replaced by
+  ``pd.read_csv(path, index_col=0, parse_dates=True)``.
+
+
 .. _whatsnew_0170.prior_deprecations:
 
 Removal of prior version deprecations/changes

pandas/core/frame.py

@@ -998,7 +998,18 @@ def from_csv(cls, path, header=0, sep=',', index_col=0,
                  parse_dates=True, encoding=None, tupleize_cols=False,
                  infer_datetime_format=False):
         """
-        Read delimited file into DataFrame
+        DEPRECATED. Please use :func:`pandas.read_csv` instead.
+
+        The preferred :func:`pandas.read_csv` only differs from this method
+        in some defaults:
+
+        - `index_col` is ``None`` instead of ``0`` (first column)
+        - `parse_dates` is ``False`` instead of ``True`` (parsing the index)
+
+        So a ``pd.DataFrame.from_csv(path)`` can be replaced by
+        ``pd.read_csv(path, index_col=0, parse_dates=True)``.
+
+        Read delimited file into DataFrame.
 
         Parameters
         ----------
@@ -1020,16 +1031,17 @@ def from_csv(cls, path, header=0, sep=',', index_col=0,
             datetime format based on the first datetime string. If the format
             can be inferred, there often will be a large parsing speed-up.
 
-        Notes
-        -----
-        Preferable to use read_table for most general purposes but from_csv
-        makes for an easy roundtrip to and from file, especially with a
-        DataFrame of time series data
+        See also
+        --------
+        pandas.read_csv
 
         Returns
         -------
         y : DataFrame
         """
+        warnings.warn("'DataFrame.from_csv' is deprecated. Use 'pandas.read_csv'"
+                      " instead (only the defaults for the 'index_col'"
+                      " and 'parse_dates' arguments differ).", FutureWarning)
         from pandas.io.parsers import read_table
         return read_table(path, header=header, sep=sep,
                           parse_dates=parse_dates, index_col=index_col,