Move sample data and add to __init__.py files

Author: Zelpuz

src/bivapp/__init__.py

@@ -0,0 +1,5 @@
+from .bivapp import (
+    BivariatePlotRaw,
+    BivariatePlotGrid,
+    BivariatePlotRawGAM,
+)

src/bivapp/bivapp.py

@@ -20,125 +20,9 @@ def to_array(self):
 
 from pygam import GAM, te  # pyGAM maintanance lapsed. Waiting for update.
 
-parent_dir = os.path.dirname(os.getcwd())
 np.random.seed(117)
 
 
-def ImportOpenairDataExample():
-    """
-    Example data from the R openair library.
-    See https://www.rdocumentation.org/packages/openair/versions/2.18-2/topics/mydata.
-    Gaseous pollutants are all in ppbv, except CO which is in ppmv.
-    Particle-phase pollutants are in ug/m3.
-    Wind speeds are in m/s.
-    Wind directions are in degrees, with 0/360 being winds from the north.
-    """
-    df = pd.read_csv(parent_dir + "/test_data/openair_test_data.csv")
-    df = df.drop(columns=["Unnamed: 0"])
-    df["date"] = pd.to_datetime(df["date"])
-    df = df.sort_index()
-
-    # Add some jitter to wind directions
-    df["wd"] = df["wd"] + np.random.uniform(-10, 10, len(df.index))
-    df["wd"] %= 360
-
-    # Add some jitter to wind speeds
-    df["ws"] = df["ws"] + np.random.uniform(-0.1, 0.1, len(df.index))
-    df.loc[df["ws"] < 0, "ws"] = 0
-
-    return df
-
-
-def ImportTorontoNO2():
-    """Imports test pollution data. The data is from 125 Resources Rd., Toronto, 2023-01-01 to 2023-12-31."""
-    # Import and rename columns
-    pol_df = pd.read_csv(parent_dir + "/test_data/NO2_2023.csv", skiprows=7)
-    pol_df.columns = [
-        "pollutant",
-        "naps_id",
-        "city",
-        "province",
-        "lat",
-        "lon",
-        "date",
-    ] + [i for i in range(1, 25)]
-
-    # Choose only Toronto 125 Resources Rd. station
-    pol_df = pol_df.loc[pol_df["naps_id"] == 60430, :]
-
-    # Wide to long
-    pol_df = pd.melt(
-        pol_df,
-        id_vars=["date"],
-        value_vars=[i for i in range(1, 25)],
-        var_name="hour",
-        value_name="no2_ppbv",
-    )
-
-    # Form and set datetime index
-    pol_df["datetime"] = pd.to_datetime(
-        pol_df["date"] + " " + (pol_df["hour"] - 1).astype(str) + ":00:00"
-    ) + pd.Timedelta("1h")
-    pol_df = pol_df.drop(columns=["date", "hour"])
-    pol_df = pol_df.set_index("datetime")
-    pol_df = pol_df.sort_index()
-
-    # Set missing values to np.nan
-    pol_df["no2_ppbv"] = pol_df["no2_ppbv"].replace(-999, np.nan)
-
-    return pol_df
-
-
-def ImportTorontoMet():
-    """Imports test meteorology. The data is from Toronto's Pearson Airport, 2023-01-01 to 2023-12-31."""
-    # Import and rename columns
-    met_df = pd.read_csv(parent_dir + "/test_data/climate-hourly.csv")
-    met_df = met_df[["UTC_DATE", "WIND_DIRECTION", "WIND_SPEED"]]
-    met_df.columns = ["datetime", "wind_dir", "wind_speed"]
-
-    # Convert from UTC to EST (pol data is in "local standard time")
-    met_df["datetime"] = pd.to_datetime(met_df["datetime"])
-    met_df["datetime"] = met_df["datetime"].dt.tz_localize("UTC").dt.tz_convert("EST")
-    met_df["datetime"] = met_df["datetime"].dt.tz_localize(None)
-    met_df = met_df.set_index("datetime")
-    met_df = met_df.sort_index()
-
-    # Drop the first row (last hour of prior year)
-    met_df = met_df.iloc[1:, :]
-
-    # Convert wind direction from tens of degrees to degrees
-    met_df["wind_dir_deg"] = met_df["wind_dir"] * 10
-    met_df = met_df.drop(columns=["wind_dir"])
-    met_df["wind_dir_deg"] = met_df["wind_dir_deg"].replace(
-        0, np.nan
-    )  # zeros are calms
-
-    # Add some jitter to wind direction
-    met_df["wind_dir_deg"] = met_df["wind_dir_deg"] + np.random.uniform(
-        -10, 10, len(met_df.index)
-    )
-    met_df["wind_dir_deg"] %= 360
-
-    # Add some jitter to wind speed and convert from km/h to m/s
-    met_df["wind_speed"] = met_df["wind_speed"] + np.random.uniform(
-        -0.5, 0.5, len(met_df.index)
-    )
-    met_df.loc[met_df["wind_speed"] < 0, "wind_speed"] = (
-        0  # the jittering could introduce negatives
-    )
-    met_df["wind_speed_m_s"] = met_df["wind_speed"] / 3.6
-    met_df = met_df.drop(columns=["wind_speed"])
-
-    return met_df
-
-
-def LoadTorontoDataExample():
-    pol_df = ImportTorontoNO2()
-    met_df = ImportTorontoMet()
-    df = pol_df.join(met_df)
-    return df
-
-
 def _makeScatterPlot(xs, ys, zs, vmin, vmax, cmap, colourbar_label, scatter_kwds):
     # Create and return the plot
     fig, ax = plt.subplots(1, 1, figsize=(8, 8), layout="constrained")

src/sample_data/__init__.py

@@ -0,0 +1,6 @@
+from ._base import (
+    ImportOpenairDataExample,
+    ImportTorontoNO2,
+    ImportTorontoMet,
+    LoadTorontoDataExample,
+)

src/sample_data/_base.py

@@ -0,0 +1,115 @@
+import pandas as pd, numpy as np
+
+def ImportOpenairDataExample():
+    """
+    Example data from the R openair library.
+    See https://www.rdocumentation.org/packages/openair/versions/2.18-2/topics/mydata.
+    Gaseous pollutants are all in ppbv, except CO which is in ppmv.
+    Particle-phase pollutants are in ug/m3.
+    Wind speeds are in m/s.
+    Wind directions are in degrees, with 0/360 being winds from the north.
+    """
+    df = pd.read_csv("/data/openair_test_data.csv")
+    df = df.drop(columns=["Unnamed: 0"])
+    df["date"] = pd.to_datetime(df["date"])
+    df = df.sort_index()
+
+    # Add some jitter to wind directions
+    df["wd"] = df["wd"] + np.random.uniform(-10, 10, len(df.index))
+    df["wd"] %= 360
+
+    # Add some jitter to wind speeds
+    df["ws"] = df["ws"] + np.random.uniform(-0.1, 0.1, len(df.index))
+    df.loc[df["ws"] < 0, "ws"] = 0
+
+    return df
+
+
+def ImportTorontoNO2():
+    """Imports test pollution data. The data is from 125 Resources Rd., Toronto, 2023-01-01 to 2023-12-31."""
+    # Import and rename columns
+    pol_df = pd.read_csv("/data/NO2_2023.csv", skiprows=7)
+    pol_df.columns = [
+        "pollutant",
+        "naps_id",
+        "city",
+        "province",
+        "lat",
+        "lon",
+        "date",
+    ] + [i for i in range(1, 25)]
+
+    # Choose only Toronto 125 Resources Rd. station
+    pol_df = pol_df.loc[pol_df["naps_id"] == 60430, :]
+
+    # Wide to long
+    pol_df = pd.melt(
+        pol_df,
+        id_vars=["date"],
+        value_vars=[i for i in range(1, 25)],
+        var_name="hour",
+        value_name="no2_ppbv",
+    )
+
+    # Form and set datetime index
+    pol_df["datetime"] = pd.to_datetime(
+        pol_df["date"] + " " + (pol_df["hour"] - 1).astype(str) + ":00:00"
+    ) + pd.Timedelta("1h")
+    pol_df = pol_df.drop(columns=["date", "hour"])
+    pol_df = pol_df.set_index("datetime")
+    pol_df = pol_df.sort_index()
+
+    # Set missing values to np.nan
+    pol_df["no2_ppbv"] = pol_df["no2_ppbv"].replace(-999, np.nan)
+
+    return pol_df
+
+
+def ImportTorontoMet():
+    """Imports test meteorology. The data is from Toronto's Pearson Airport, 2023-01-01 to 2023-12-31."""
+    # Import and rename columns
+    met_df = pd.read_csv("/data/climate-hourly.csv")
+    met_df = met_df[["UTC_DATE", "WIND_DIRECTION", "WIND_SPEED"]]
+    met_df.columns = ["datetime", "wind_dir", "wind_speed"]
+
+    # Convert from UTC to EST (pol data is in "local standard time")
+    met_df["datetime"] = pd.to_datetime(met_df["datetime"])
+    met_df["datetime"] = met_df["datetime"].dt.tz_localize("UTC").dt.tz_convert("EST")
+    met_df["datetime"] = met_df["datetime"].dt.tz_localize(None)
+    met_df = met_df.set_index("datetime")
+    met_df = met_df.sort_index()
+
+    # Drop the first row (last hour of prior year)
+    met_df = met_df.iloc[1:, :]
+
+    # Convert wind direction from tens of degrees to degrees
+    met_df["wind_dir_deg"] = met_df["wind_dir"] * 10
+    met_df = met_df.drop(columns=["wind_dir"])
+    met_df["wind_dir_deg"] = met_df["wind_dir_deg"].replace(
+        0, np.nan
+    )  # zeros are calms
+
+    # Add some jitter to wind direction
+    met_df["wind_dir_deg"] = met_df["wind_dir_deg"] + np.random.uniform(
+        -10, 10, len(met_df.index)
+    )
+    met_df["wind_dir_deg"] %= 360
+
+    # Add some jitter to wind speed and convert from km/h to m/s
+    met_df["wind_speed"] = met_df["wind_speed"] + np.random.uniform(
+        -0.5, 0.5, len(met_df.index)
+    )
+    met_df.loc[met_df["wind_speed"] < 0, "wind_speed"] = (
+        0  # the jittering could introduce negatives
+    )
+    met_df["wind_speed_m_s"] = met_df["wind_speed"] / 3.6
+    met_df = met_df.drop(columns=["wind_speed"])
+
+    return met_df
+
+
+def LoadTorontoDataExample():
+    pol_df = ImportTorontoNO2()
+    met_df = ImportTorontoMet()
+    df = pol_df.join(met_df)
+    return df