Implement get_day_of_year, tests (#19555)

Author: jbrockmendel

pandas/_libs/tslibs/ccalendar.pxd

@@ -10,3 +10,4 @@ cdef int dayofweek(int y, int m, int m) nogil
 cdef bint is_leapyear(int64_t year) nogil
 cpdef int32_t get_days_in_month(int year, Py_ssize_t month) nogil
 cpdef int32_t get_week_of_year(int year, int month, int day) nogil
+cpdef int32_t get_day_of_year(int year, int month, int day) nogil

pandas/_libs/tslibs/ccalendar.pyx

@@ -142,17 +142,13 @@ cpdef int32_t get_week_of_year(int year, int month, int day) nogil:
     Assumes the inputs describe a valid date.
     """
     cdef:
-        bint isleap, isleap_prev
-        int32_t mo_off
+        bint isleap
         int32_t doy, dow
         int woy
 
     isleap = is_leapyear(year)
-    isleap_prev = is_leapyear(year - 1)
-
-    mo_off = _month_offset[isleap * 13 + month - 1]
 
-    doy = mo_off + day
+    doy = get_day_of_year(year, month, day)
     dow = dayofweek(year, month, day)
 
     # estimate
@@ -162,7 +158,7 @@ cpdef int32_t get_week_of_year(int year, int month, int day) nogil:
 
     # verify
     if woy < 0:
-        if (woy > -2) or (woy == -2 and isleap_prev):
+        if (woy > -2) or (woy == -2 and is_leapyear(year - 1)):
             woy = 53
         else:
             woy = 52
@@ -171,3 +167,36 @@ cpdef int32_t get_week_of_year(int year, int month, int day) nogil:
             woy = 1
 
     return woy
+
+
+@cython.wraparound(False)
+@cython.boundscheck(False)
+cpdef int32_t get_day_of_year(int year, int month, int day) nogil:
+    """Return the ordinal day-of-year for the given day.
+
+    Parameters
+    ----------
+    year : int
+    month : int
+    day : int
+
+    Returns
+    -------
+    day_of_year : int32_t
+
+    Notes
+    -----
+    Assumes the inputs describe a valid date.
+    """
+    cdef:
+        bint isleap
+        int32_t mo_off
+        int32_t doy, dow
+        int woy
+
+    isleap = is_leapyear(year)
+
+    mo_off = _month_offset[isleap * 13 + month - 1]
+
+    day_of_year = mo_off + day
+    return day_of_year

pandas/_libs/tslibs/fields.pyx

@@ -15,7 +15,7 @@ cnp.import_array()
 
 
 from ccalendar cimport (get_days_in_month, is_leapyear, dayofweek,
-                        get_week_of_year)
+                        get_week_of_year, get_day_of_year)
 from np_datetime cimport (pandas_datetimestruct, pandas_timedeltastruct,
                           dt64_to_dtstruct, td64_to_tdstruct)
 from nattype cimport NPY_NAT
@@ -374,15 +374,7 @@ def get_date_field(ndarray[int64_t] dtindex, object field):
     cdef:
         Py_ssize_t i, count = 0
         ndarray[int32_t] out
-        ndarray[int32_t, ndim=2] _month_offset
-        int isleap, isleap_prev
         pandas_datetimestruct dts
-        int mo_off, doy, dow
-
-    _month_offset = np.array(
-        [[0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365],
-         [0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366]],
-        dtype=np.int32 )
 
     count = len(dtindex)
     out = np.empty(count, dtype='i4')
@@ -482,8 +474,7 @@ def get_date_field(ndarray[int64_t] dtindex, object field):
                     continue
 
                 dt64_to_dtstruct(dtindex[i], &dts)
-                isleap = is_leapyear(dts.year)
-                out[i] = _month_offset[isleap, dts.month -1] + dts.day
+                out[i] = get_day_of_year(dts.year, dts.month, dts.day)
         return out
 
     elif field == 'dow':

pandas/_libs/tslibs/period.pyx

@@ -22,7 +22,7 @@ from cpython.datetime cimport PyDateTime_Check, PyDateTime_IMPORT
 PyDateTime_IMPORT
 
 from np_datetime cimport (pandas_datetimestruct, dtstruct_to_dt64,
-                          dt64_to_dtstruct, is_leapyear)
+                          dt64_to_dtstruct)
 
 cimport util
 from util cimport is_period_object, is_string_object, INT32_MIN
@@ -34,11 +34,12 @@ from timezones cimport is_utc, is_tzlocal, get_utcoffset, get_dst_info
 from timedeltas cimport delta_to_nanoseconds
 
 from ccalendar import MONTH_NUMBERS
+from ccalendar cimport is_leapyear
 from frequencies cimport (get_freq_code, get_base_alias,
                           get_to_timestamp_base, get_freq_str,
                           get_rule_month)
 from parsing import parse_time_string, NAT_SENTINEL
-from resolution import resolution, Resolution
+from resolution import Resolution
 from nattype import nat_strings, NaT, iNaT
 from nattype cimport _nat_scalar_rules, NPY_NAT
 

pandas/_libs/tslibs/timestamps.pyx

@@ -29,8 +29,7 @@ from nattype import NaT
 from nattype cimport NPY_NAT
 from np_datetime import OutOfBoundsDatetime
 from np_datetime cimport (reverse_ops, cmp_scalar, check_dts_bounds,
-                          pandas_datetimestruct, dt64_to_dtstruct,
-                          is_leapyear)
+                          pandas_datetimestruct, dt64_to_dtstruct)
 from timedeltas import Timedelta
 from timedeltas cimport delta_to_nanoseconds
 from timezones cimport (
@@ -291,14 +290,6 @@ cdef class _Timestamp(datetime):
             val = tz_convert_single(self.value, 'UTC', self.tz)
         return val
 
-    cpdef int _get_field(self, field):
-        cdef:
-            int64_t val
-            ndarray[int32_t] out
-        val = self._maybe_convert_value_to_local()
-        out = get_date_field(np.array([val], dtype=np.int64), field)
-        return int(out[0])
-
     cpdef bint _get_start_end_field(self, str field):
         cdef:
             int64_t val
@@ -695,14 +686,11 @@ class Timestamp(_Timestamp):
 
     @property
     def dayofyear(self):
-        return self._get_field('doy')
+        return ccalendar.get_day_of_year(self.year, self.month, self.day)
 
     @property
     def week(self):
-        if self.freq is None:
-            # fastpath for non-business
-            return ccalendar.get_week_of_year(self.year, self.month, self.day)
-        return self._get_field('woy')
+        return ccalendar.get_week_of_year(self.year, self.month, self.day)
 
     weekofyear = week
 
@@ -764,7 +752,7 @@ class Timestamp(_Timestamp):
 
     @property
     def is_leap_year(self):
-        return bool(is_leapyear(self.year))
+        return bool(ccalendar.is_leapyear(self.year))
 
     def tz_localize(self, tz, ambiguous='raise', errors='raise'):
         """

pandas/core/indexes/datetimes.py

@@ -55,7 +55,7 @@
 from pandas._libs import (lib, index as libindex, tslib as libts,
                           join as libjoin, Timestamp)
 from pandas._libs.tslibs import (timezones, conversion, fields, parsing,
-                                 period as libperiod)
+                                 resolution as libresolution)
 
 # -------- some conversion wrapper functions
 
@@ -1795,7 +1795,7 @@ def is_normalized(self):
 
     @cache_readonly
     def _resolution(self):
-        return libperiod.resolution(self.asi8, self.tz)
+        return libresolution.resolution(self.asi8, self.tz)
 
     def insert(self, loc, item):
         """

pandas/tests/tslibs/__init__.py

@@ -0,0 +1,18 @@
+# -*- coding: utf-8 -*-
+from datetime import datetime
+
+import numpy as np
+
+from pandas._libs.tslibs import ccalendar
+
+
+def test_get_day_of_year():
+    assert ccalendar.get_day_of_year(2001, 3, 1) == 60
+    assert ccalendar.get_day_of_year(2004, 3, 1) == 61
+    assert ccalendar.get_day_of_year(1907, 12, 31) == 365
+    assert ccalendar.get_day_of_year(2004, 12, 31) == 366
+
+    dt = datetime.fromordinal(1 + np.random.randint(365 * 4000))
+    result = ccalendar.get_day_of_year(dt.year, dt.month, dt.day)
+    expected = (dt - dt.replace(month=1, day=1)).days + 1
+    assert result == expected

pandas/tests/tslibs/test_ccalendar.py

@@ -515,6 +515,7 @@ def pxd(name):
         'pyxfile': '_libs/tslibs/period',
         'pxdfiles': ['_libs/src/util',
                      '_libs/missing',
+                     '_libs/tslibs/ccalendar',
                      '_libs/tslibs/timedeltas',
                      '_libs/tslibs/timezones',
                      '_libs/tslibs/nattype'],