Task 10

README.md

@@ -1,8 +1,56 @@
-# learning-by-doing
-
-a set of tasks for people to learn basic software and data science skills by completing the tasks
-
-
-## License
-
-Except as otherwise noted, the tutorial content of this `astropgh/learning-by-doing` is licensed under the [Creative Commons Attribution-ShareAlike 4.0 International License](http://creativecommons.org/licenses/by-sa/4.0/) [![CC BY-SA 4.0](https://i.creativecommons.org/l/by-sa/4.0/80x15.png)](http://creativecommons.org/licenses/by-sa/4.0/), and the code samples are licensed under the [MIT License](https://opensource.org/licenses/MIT). 
+# Task 4: [database]  Preparing for data scraping: design a data model for top baby names
+## Background
+Before we start to scrape the top baby names from the webpage, we need to design
+a data model that we will use to store the data.
+The term "data model" has different meanings in different contexts.
+We can ask what kind of object the data will be stored in.
+A python list? A python dictionary? A pandas data frame?
+For a given type, we can further ask how the data is stored.
+For example, if we store the data in a pandas data frame, we can ask what
+are the columns and rows.
+Let's look at some examples.
+The original webpage store the names as a table, with columns being
+`year`, `female_rank1`, `female_rank2`, `male_rank1`, `male_rank2`..., and
+each row corresponds to one single year.
+A more extreme example would be storing the names as a sequence (say a python list),
+the content of the sequence will be the names, while the indices of the sequence encode
+year, ranking, and gender altogether. A possible way to encode the information is
+```python
+year = 2017 - index // 10
+rank = index % 5 + 1
+gender = 'female' if index % 10 < 5 else 'male'
+```
+While this data model preserves all the information, it is unlikely that this
+model will be very convenient when it comes to data exploration.
+Yet another totally different data model is to group the data by names.
+Let's say we'll store the data in a python dictionary. A possible way is:
+```python
+{
+    'Emma':{
+        'gender': 'female',
+        'years_ranked_1': [2017, 2016, 2015, 2014, ...],
+        'years_ranked_2': [2013, 2012, 2009, ...],
+        'years_ranked_3': [...],
+    },
+    'Noah':{
+        ...,
+    },
+    ...,
+}
+```
+Note that the form (object) that the data is stored and how the data is structured
+are two different things. (*Food for thoughts: why? can you give an example?*)
+Clearly, the choice of data model heavily depends on the questions that we would
+like to answer with the data.
+If the amount of data is very large, we will also need to consider the avabilable
+computing resources like memory usage and I/O speed when designing the data model.
+For now, we don't yet need to worry about the limitation due to computing resources.
+## Task
+Try to come up with a data model that is good for answering each of the following questions.
+Think about the code you'll need to write to interact with the data model to answer
+these questions.
+1. Which years Emma is the most chosen names?
+2. Which name had been the most chosen name for the longest consecutive years?
+3. How many unique male names have be on top 5 between years 1980 and 2000?
+4. Are there more unique male names or more unique female names that are on top 5?
+5. What is the distribution of the numbers of consecutive years that a male name remains the most chosen name?