Add solution package

Author: Tania Allard

.gitignore/.gitignore

@@ -102,3 +102,4 @@ ENV/
 .mypy_cache/
 
 /data
+\.DS_Store

03_ProcessData.ipynb

@@ -157,6 +157,24 @@
     "```\n"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {
+    "slideshow": {
+     "slide_type": "subslide"
+    }
+   },
+   "source": [
+    "# Packaging\n",
+    "\n",
+    "Since are using a modular approach we need to ensure we can call our functions from a **run all script**.\n",
+    "\n",
+    "From the shell: \n",
+    "```\n",
+    "$ touch src/__init__/py\n",
+    "```"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 1,

solutions/scripts/01_runall-wine-analysis.py

@@ -0,0 +1,41 @@
+#!/usr/bin/env python
+
+"""
+Script to create all the results from our
+amazing wine study
+"""
+
+import numpy as np
+import pandas as pd
+import matplotlib.pyplot as plt
+import importlib
+
+# imports from our package
+# Since we have digits at the start of the modules we
+# will use dynamic imports
+subset = importlib.import_module('.data.01_subset-data-GBP', 'scripts')
+plotwines = importlib.import_module('.visualization.02_visualize-wines', 'scripts')
+country_sub = importlib.import_module('.data.03_country-subset', 'scripts')
+
+# ------------------------------------------------------------------------
+# Declare variables
+# ------------------------------------------------------------------------
+
+# Set raw data path
+raw_data = "data/raw/winemag-data-130k-v2.csv"
+interim_data = "data/interim/2018-05-02-winemag_priceGBP.csv"
+
+# Set country
+country = "Chile"
+
+
+# ------------------------------------------------------------------------
+# Perform analysis
+# ------------------------------------------------------------------------
+
+if __name__ == '__main__':
+    subset_file = subset.process_data_GBP(raw_data)
+    print(subset_file)
+    plotwines.create_plots(interim_data)
+    country_file = country_sub.get_country(subset_file, country)
+    print(country_file)

solutions/scripts/__init__.py

@@ -0,0 +1,2 @@
+from . import data
+from . import visualization

solutions/scripts/__pycache__/01_runall-wine-analysis.cpython-36.pyc

@@ -0,0 +1,46 @@
+#!/usr/bin/env python
+
+import sys
+import datetime
+
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+
+
+def process_data_GBP(filename):
+    """
+    Get only the needed subset from the data.
+    Args:
+    filename: str
+        Path to the filename containing the wine data
+
+    Returns:
+
+    data_path: st
+        Path to the created data set
+    """
+
+    # Load table
+    wine = pd.read_csv(filename)
+
+    # Subset of data to keep
+    wine_keep = wine.loc[:,['country', 'designation', 'points', 'price']]
+
+    # Add column with prices in GBP
+    wine_keep['price_GBP'] = wine_keep['price'].apply(lambda x : x * 1.2)
+
+    # Constructing the fname
+    today = datetime.datetime.today().strftime('%Y-%m-%d')
+    fname = f'data/interim/{today}-winemag_priceGBP.csv'
+
+    # Saving the csv
+    wine_keep.to_csv(fname)
+
+    return(fname)
+
+
+if __name__ == '__main__':
+    filename = sys.argv[1]
+    print(filename)
+    print(process_data_GBP(filename))

solutions/scripts/data/.gitkeep

@@ -0,0 +1,51 @@
+#!/usr/bin/env python
+
+
+import sys
+import datetime
+
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+
+
+def get_country(filename, country):
+    """
+    Do a simple analysis per country
+    Args:
+    filename: str
+        Path to the filename containing the wine data
+    country: str
+        Country to be used to subset
+
+    Returns:
+
+    data_path: st
+        Path to the created data set
+    """
+
+    # Load table
+    wine = pd.read_csv(filename)
+
+    # Use the country name to subset data
+    subset_country = wine[wine['country'] == country ].copy()
+
+    # Subset the
+
+    # Constructing the fname
+    today = datetime.datetime.today().strftime('%Y-%m-%d')
+    fname = f'data/processed/{today}-winemag_{country}.csv'
+
+    # Saving the csv
+    subset_country.to_csv(fname)
+
+    return(fname)
+
+
+if __name__ == '__main__':
+    filename = sys.argv[1]
+    country = sys.argv[2]
+    print(f'Subsetting: {filename}')
+    print(f'Country searched: {country}')
+
+    print(get_country(filename, country))