Merge pull request #63 from Stanford-NavLab/derek/init-simplify

Derek/init simplify

Author: kanhereashwin

gnss_lib_py/parsers/android.py

@@ -33,12 +33,7 @@ def __init__(self, input_path):
             Loaded measurements with consistent column names
         """
 
-        pd_df = pd.read_csv(input_path)
-        col_map = self._column_map()
-        pd_df.rename(columns=col_map, inplace=True)
-
-        super().__init__(pandas_df=pd_df)
-        self.postprocess()
+        super().__init__(csv_path=input_path)
 
     def postprocess(self):
         """Android derived specific postprocessing
@@ -57,15 +52,15 @@ def postprocess(self):
         self['corr_pr_m'] = pr_corrected
 
     @staticmethod
-    def _column_map():
+    def _row_map():
         """Map of column names from loaded to gnss_lib_py standard
 
         Returns
         -------
-        col_map : Dict
+        row_map : Dict
             Dictionary of the form {old_name : new_name}
         """
-        col_map = {'collectionName' : 'trace_name',
+        row_map = {'collectionName' : 'trace_name',
                    'phoneName' : 'rx_name',
                    'constellationType' : 'gnss_id',
                    'svid' : 'sv_id',
@@ -84,7 +79,7 @@ def _column_map():
                    'ionoDelayM' : 'iono_delay_m',
                    'tropoDelayM' : 'tropo_delay_m',
                    }
-        return col_map
+        return row_map
 
 
 class AndroidRawImu(NavData):
@@ -137,19 +132,18 @@ def preprocess(self, input_path):
         #NOTE: Assuming pandas index corresponds to measurements order
         #NOTE: Override times of gyro measurements with corresponding
         # accel times
-        measurements.rename(columns=self._column_map(), inplace=True)
         return measurements
 
     @staticmethod
-    def _column_map():
-        col_map = {'AccelXMps2' : 'acc_x_mps2',
+    def _row_map():
+        row_map = {'AccelXMps2' : 'acc_x_mps2',
                    'AccelYMps2' : 'acc_y_mps2',
                    'AccelZMps2' : 'acc_z_mps2',
                    'GyroXRadPerSec' : 'ang_vel_x_radps',
                    'GyroYRadPerSec' : 'ang_vel_y_radps',
                    'GyroZRadPerSec' : 'ang_vel_z_radps',
                    }
-        return col_map
+        return row_map
 
 
 class AndroidRawFixes(NavData):

gnss_lib_py/parsers/navdata.py

@@ -51,6 +51,8 @@ def __init__(self, csv_path=None, pandas_df=None, numpy_array=None,
         else:
             self.build_navdata()
 
+        self.rename(self._row_map())
+
         self.postprocess()
 
     def postprocess(self):
@@ -125,6 +127,23 @@ def from_numpy_array(self, numpy_array):
         for row_num in range(numpy_array.shape[0]):
             self.__setitem__(str(row_num), numpy_array[row_num,:])
 
+    @staticmethod
+    def _row_map():
+        """Map of column names from loaded to gnss_lib_py standard
+
+        Initializes as an emptry dictionary, must be reimplemented for
+        custom parsers.
+
+        Returns
+        -------
+        row_map : Dict
+            Dictionary of the form {old_name : new_name}
+        """
+
+        row_map = {}
+
+        return row_map
+
     def pandas_df(self):
         """Return pandas DataFrame equivalent to class
 
@@ -296,6 +315,7 @@ def __setitem__(self, key_idx, newvalue):
             #Creating an entire new row
             if isinstance(newvalue, np.ndarray) and newvalue.dtype==object:
                 # Adding string values
+                # print("\n",key_idx,"\n",newvalue)
                 self.fillna(newvalue)
                 new_str_vals = len(np.unique(newvalue))*np.ones(np.shape(newvalue),
                                     dtype=self.arr_dtype)
@@ -308,6 +328,7 @@ def __setitem__(self, key_idx, newvalue):
                     self.array = np.vstack((self.array, np.reshape(new_str_vals, [1, -1])))
                 self.map[key_idx] = self.shape[0]-1
             else:
+                # print("\n",key_idx,"\n")#,newvalue)
                 if not isinstance(newvalue, int) and not isinstance(newvalue, float):
                     assert not isinstance(np.asarray(newvalue)[0], str), \
                             "Cannot set a row with list of strings, please use np.ndarray with dtype=object"
@@ -665,7 +686,7 @@ def rename(self, mapper):
         for key, value in mapper.items():
             if not isinstance(value, str):
                 raise TypeError("Column names must be strings")
-            if key not in self.map.keys():
+            if key not in self.map:
                 raise KeyError("'" + str(key) \
                                + "' key doesn't exist in NavData class")
 

notebooks/tutorials/parsers.ipynb

@@ -193,7 +193,7 @@
    "source": [
     "One thing that we need to do to make use of the common functionality of `gnss_lib_py` is to standardize the names of our variables. See the [Standard Naming Conventions](https://gnss-lib-py.readthedocs.io/en/latest/reference/reference.html#standard-naming-conventions) section in the Reference tab of the documentation for the list of standardized names.\n",
     "\n",
-    "In our case, we will convert `mySatId` to `sv_id` and `myPseudorange` to `raw_pr_m`. We make these conversions by creating the `_col_map` function and calling the `rename` function in our postprocessing step."
+    "In our case, we will convert `mySatId` to `sv_id` and `myPseudorange` to `raw_pr_m`. We make these conversions by simply updating the `_row_map` function."
    ]
   },
   {
@@ -203,23 +203,20 @@
    "source": [
     "```python\n",
     "\n",
-    "# rename rows\n",
-    "col_map = self._column_map()\n",
-    "self.rename(col_map)\n",
     "\n",
     "@staticmethod\n",
-    "def _column_map():\n",
+    "def _row_map():\n",
     "    \"\"\"Map of column names from loaded to gnss_lib_py standard\n",
     "\n",
     "    Returns\n",
     "    -------\n",
-    "    col_map : Dict\n",
+    "    row_map : Dict\n",
     "        Dictionary of the form {old_name : new_name}\n",
     "    \"\"\"\n",
-    "    col_map = {'mySatId' : 'sv_id',\n",
+    "    row_map = {'mySatId' : 'sv_id',\n",
     "               'myPseudorange' : 'raw_pr_m',\n",
     "               }\n",
-    "    return col_map\n",
+    "    return row_map\n",
     "```"
    ]
   },
@@ -284,11 +281,7 @@
     "        \"\"\"MyReceiver specific postprocessing\n",
     "\n",
     "        \"\"\"\n",
-    "        \n",
-    "        # rename rows\n",
-    "        col_map = self._column_map()\n",
-    "        self.rename(col_map)\n",
-    "        \n",
+    "\n",
     "        # correct pseudorange\n",
     "        self['corr_pr_m'] = self['raw_pr_m'] + 100.\n",
     "        \n",
@@ -297,18 +290,18 @@
     "        \n",
     "\n",
     "    @staticmethod\n",
-    "    def _column_map():\n",
+    "    def _row_map():\n",
     "        \"\"\"Map of column names from loaded to gnss_lib_py standard\n",
     "\n",
     "        Returns\n",
     "        -------\n",
-    "        col_map : Dict\n",
+    "        row_map : Dict\n",
     "            Dictionary of the form {old_name : new_name}\n",
     "        \"\"\"\n",
-    "        col_map = {'mySatId' : 'sv_id',\n",
+    "        row_map = {'mySatId' : 'sv_id',\n",
     "                   'myPseudorange' : 'raw_pr_m',\n",
     "                   }\n",
-    "        return col_map"
+    "        return row_map"
    ]
   },
   {

tests/parsers/test_android.py

@@ -107,16 +107,16 @@ def fixture_pd_df(derived_path):
     return derived_df
 
 
-@pytest.fixture(name="derived_col_map")
-def fixture_inverse_col_map():
+@pytest.fixture(name="derived_row_map")
+def fixture_inverse_row_map():
     """Map from standard names to derived column names
 
     Returns
     -------
-    inverse_col_map : Dict
+    inverse_row_map : Dict
         Column names for inverse map of form {standard_name : derived_name}
     """
-    inverse_col_map = {'trace_name' : 'collectionName',
+    inverse_row_map = {'trace_name' : 'collectionName',
                        'rx_name' : 'phoneName',
                        'gnss_id' : 'constellationType',
                        'sv_id' : 'svid',
@@ -135,7 +135,7 @@ def fixture_inverse_col_map():
                        'iono_delay_m' : 'ionoDelayM',
                        'tropo_delay_m' : 'tropoDelayM',
                     }
-    return inverse_col_map
+    return inverse_row_map
 
 
 @pytest.fixture(name="derived")
@@ -156,7 +156,7 @@ def fixture_load_derived(derived_path):
     return derived
 
 
-def test_derived_df_equivalence(derived, pd_df, derived_col_map):
+def test_derived_df_equivalence(derived, pd_df, derived_row_map):
     """Test if naive dataframe and AndroidDerived contain same data
 
     Parameters
@@ -165,15 +165,17 @@ def test_derived_df_equivalence(derived, pd_df, derived_col_map):
         Instance of AndroidDerived for testing
     pd_df : pytest.fixture
         pd.DataFrame for testing measurements
-    derived_col_map : pytest.fixture
+    derived_row_map : pytest.fixture
         Column map to convert standard to original derived column names
 
     """
     # Also tests if strings are being converted back correctly
     measure_df = derived.pandas_df()
-    measure_df.rename(columns=derived_col_map, inplace=True)
+    measure_df.rename(columns=derived_row_map, inplace=True)
     measure_df = measure_df.drop(columns='corr_pr_m')
-    pd.testing.assert_frame_equal(pd_df, measure_df, check_dtype=False)
+    pd.testing.assert_frame_equal(pd_df.sort_index(axis=1),
+                                  measure_df.sort_index(axis=1),
+                                  check_dtype=False, check_names=True)
 
 
 @pytest.mark.parametrize('row_name, index, value',
@@ -328,10 +330,10 @@ def test_csv_equivalence(android_raw_path, root_path, file_type):
     csv_loc = make_csv(android_raw_path, output_directory, file_type)
     test_df = pd.read_csv(csv_loc)
     test_measure = AndroidRawImu(android_raw_path)
-    col_map = test_measure._column_map()
+    row_map = test_measure._row_map()
     for col_name in test_df.columns:
-        if col_name in col_map:
-            row_idx = col_map[col_name]
+        if col_name in row_map:
+            row_idx = row_map[col_name]
         else:
             row_idx = col_name
         if col_name in no_check or col_name :

tests/parsers/test_navdata.py

@@ -958,7 +958,5 @@ def test_col_looping(csv_simple):
     for idx, col in enumerate(data):
         col_df = col.pandas_df().reset_index(drop=True)
         expected_df = compare_df.iloc[[idx], :].reset_index(drop=True)
-        print(col_df)
-        print(expected_df)
         pd.testing.assert_frame_equal(col_df, expected_df,
                                       check_index_type=False)

tests/utils/test_visualizations.py

@@ -113,8 +113,8 @@ def test_plot_metrics(derived):
                  "tropo_delay_m",
                  ]
 
-    for r_idx, row in enumerate(derived.rows):
-        if not derived.str_bool[r_idx]:
+    for row in derived.rows:
+        if not derived.str_bool[derived.map[row]]:
             if row in test_rows:
                 fig = viz.plot_metric(derived, row, save=False)
                 close_figures(fig)
@@ -129,8 +129,8 @@ def test_plot_metrics(derived):
         viz.plot_metric(derived, "raw_pr_m", save=True, prefix=1)
     assert "Prefix" in str(excinfo.value)
 
-    for r_idx, row in enumerate(derived.rows):
-        if not derived.str_bool[r_idx]:
+    for row in derived.rows:
+        if not derived.str_bool[derived.map[row]]:
             if row in test_rows:
                 fig = viz.plot_metric(derived, "raw_pr_m", row, save=False)
                 close_figures(fig)
@@ -167,8 +167,8 @@ def test_plot_metrics_by_constellation(derived):
                  "tropo_delay_m",
                  ]
 
-    for r_idx, row in enumerate(derived.rows):
-        if not derived.str_bool[r_idx]:
+    for row in derived.rows:
+        if not derived.str_bool[derived.map[row]]:
             if row in test_rows:
                 fig = viz.plot_metric_by_constellation(derived, row, save=False)
                 close_figures(fig)