Project 1: Combining Data from Multiple Sources
Introduction
Finance research often requires assembling a data set from different sources. In many cases, the sourcedata is not formatted such that it can be imported easily into Python for analysis. This assessment helpsyou develop the foundational skills of data acquisition, cleaning, and merging. You will combine stock priceinformation distributed across many files and produce the output as a JSON file. We ask you to write generalcode, such that it can adapt easily to different file formats.Writing general code can be a daunting exercise at first. However, it is in your best interest to practicewriting functions that can adapt to different configurations. Doing so makes your code more robust, easierto maintain, and upgradeable.
To help you, we have provided a Python scaffold. This file, called zid_project1.py contains all the functionsneeded for this project. Each function has a detailed docstring describing what the function does, the inputparameters, and the object it returns.The remainder of this document provide information on:
- The data files you will receive.
- Instructions to set up your development environment in PyCharm.
- Detailed step-by-step instructions required to complete the assessment. Please follow these instructionsclosely. Our ability to evaluate your work requires that you do so.
You should develop your code within PyCharm. Submission, however, will be through Ed. You will only
need to copy your zid_project1.py file into Ed. Unlike the code challenges you have done so far, Ed will
not provide you any feedback on your code. You can still submit multiple times before the deadline – Only
your final submission on Ed will be marked.
The Source Files
All required files are included in a zip folder with the following structure:
project1/
| project_desc.pdf
| README.txt
| TICEKRS.txt
| zid_project1.py <-- This is the only file you need to submit
|___data/
| | <tic>_prc.dat <-- Several files of this type
- zid_project1.py contains the functions you will use in this assessment. Some of the functions are
already written, while others you will need to write. Please see the instructions below for more
information.
- README.txt contains information about how the stock data is stored in the .dat files. Please use
the information contained the README.txt file you received (there are different versions for different
students).
- TICKERS.txt contains a list of tickers and their corresponding exchanges, one per line. These tickers
and exchanges may be in upper or lower case.
1• project_desc.pdf is a PDF version of this document.
- data is a sub-folder containing all the data you will need to complete this assessment. Inside this folder
you will find many files. Each <tic>_prc.dat contains stock price data for the ticker <tic>. Every
ticker in TICKERS.txt will have a corresponding “.dat” file. However, you may have more “.dat” files
than tickers in the TICKERS.txt file. In other words, you may have more “.dat” files than you will need
(but not fewer).
Instructions
To set up your PyCharm for this assessment, please follow the following steps:
- Unzip the contents of the zip file onto your computer.
- Copy/move the entire project1 folder into your PyCharm toolkit project folder. Afterwards, your
toolkit folder will look like:
toolkit/ <-- Project folder
| toolkit_config.py <-- Already created
| ...
|___project1/ <-- Contents of the zipped folder
| | project_desc.pdf
| | README.txt
| | TICEKRS.txt
| | zid_project1.py
| |
| |___data/
| | | <tic>_prc.dat
- Complete the user-written functions in zid_project1.py. See the step-by-step instructions in Com
pleting the Code Scaffold below.
- After you have completed the zid_project1.py module in PyCharm, copy and paste the entire con
tents of this module to Ed. This is the only file you will need to submit to complete this assessment.
- Press “Submit” to submit your project. Your project will not be submitted until you do so.
Completing the Code Scaffold
After setting up your PyCharm development environment with the project files (see instructions above),
modify the zid_project1.py module by taking the the steps described below, in sequence. The completed
code will produce a JSON containing the combined contents of several files. You do not need to submit this
JSON file.
Step 1: Set the location of files and folders (3 marks)
Open the zid_project1.py module in PyCharm.
Set the correct expressions for the constants ROOTDIR, DATDIR, and TICPATH as described below. Importantly,
you should not include forward slashes or backslashes when defining these variables (so no “C:\User. . .”,
etc. . . ). Instead, you should use the appropriate methods from the os module.
2• The ROOTDIR variable combines the base location of the toolkit project folder (which is already
specified in your toolkit_config.py module) and the project1 package.
- The DATDIR variable combines the location in ROOTDIR above and the data sub-folder. Note that this
is a different variable than the DATADIR included in your toolkit_config.py module. The name of
the DATDIR variable is a combination of “DAT” and “DIR” (not “DATA” + “DIR”), and it points to
a different location in your computer.
- The TICPATH variable combines the location in ROOTDIR and the name of the file with the tickers
(TICKERS.txt).
Again, all these paths should be created using the appropriate methods from the os module. If you include
any forward or backward slashes in the definition of these variables your code will only run in your computer.
A big part of this assessment is to make sure your code is portable.
3The diagram below presents the relation between these variables and their locations. Again, do not use full
paths (like “C:\Users. . .”) to create these variables.
toolkit/
| toolkit_config.py <-- You already created this module
| ...
|___project1/ <-- `ROOTDIR` variable poitns to this folder
| | ...
| | TICEKRS.txt <-- `TICPATH` variables |points to this file
| | ...
| | zid_project1.py
| |___data/ <-- `DATDIR` variable |points to this folder
| | | ...
Note: All you have to do for this part is to replace the strings "<COMPLETE THIS PART>" with the appropriate
expressions.
Step 2: Set the variables describing the format of the source data (2 marks)
This part of the project is very important! Make sure you follow these instructions closely.
Before we start, open one of the “.dat” files in PyCharm. To do that, just navigate to the
toolkit/project1/data/ folder (inside PyCharm) and double click on one of the files. You will
notice the following:
- There are no column headers. Every line in this file (including the first one) contains data.
- There is no “separator” between columns (e.g., columns are not separated by comma, tabs, etc. . .
). Instead, columns have a fixed width – For example, the first 10 characters belong to column 1, the
next 8 to column 2, etc. . .
This means we have to create a function to split the lines into columns, so that each “value” is assigned to its
correct “data field”. The first step is to set the correct expressions for the variables COLUMNS and COLWIDTHS.
- The COLUMNS variable must be a list, where each element represents a source column name in the
README.txt file. The order of the elements in this list must match the order of the columns in the
README.txt file. For instance, suppose you have the following information in the README.txt file:
# ----------------------------------------------------------------------------
# Column information
# ----------------------------------------------------------------------------
Close:
column position: 1
width: 14
Date:
column position: 2
width: 11
In this case, you must set COLUMNS = ['Close', 'Date'].
- The COLWIDTHS variable must be a dictionary. Each key is a column name in COLUMNS. Each value is
the width of this column in the README.txt file. In the example above, you would set COLSWIDTH =
{'Close': 14, 'Date': 11}.
4Step 3: Complete the get_tics function (15 marks)
Complete the indicated part of the function get_tics. This function reads a file with tickers and returns a
list with formatted tickers. Make sure the function works with the given pth variable and not the constant
TICPATH (i.e., there should be no reference to the TICPATH constant inside this function). We will test your
code using different files. Using TICPATH instead of pth inside get_tics means that your function always
returns the same tickers instead of adapting to different possible ticker lists.
Your function must be consistent with the docstring provided. In particular, please make sure the body of
the function is consistent with the “Parameters” and “Returns” sections of the docstring. The only exception
is the optional suggestions provided in the “Hints” section, which you do not need to follow.
Your module also includes a test function called _test_get_tics. After you finish creating the get_tics
function, it is a good idea to run the _test_get_tics function and look at the output. That should give
you a good indication if your function is performing as expected. You can uncomment the relevant part of
the if __name__ ... code block to run this test function. Like all other test functions provided, you can
modify or delete these functions – they will not be marked.
Step 4: Complete the read_dat function (5 marks)
Complete the indicated part of the function read_dat. This function reads a stock price data file for a given
ticker and returns its contents as a list of lines. Please make sure the body of the function is consistent with
the “Parameters” and “Returns” sections of the docstring. You may choose to follow the proposed steps in
the “Hints” section but that is optional.
Remember not to use literals with full paths like “C:\Users. . .” inside the body of the function (or anywhere
in the module). You can use the constants you created in step 1 above (e.g., DATDIR) and methods from
the os module to create paths. You can use the corresponding “test” function _test_read_dat to test this
function once its completed.
Step 5: Complete the line_to_dict function (10 marks)
Complete the indicated part of the function line_to_dict. The same instructions provided for the get_tics
and read_dat functions above apply to this function as well.
Step 6: Complete the verify_tickers function (10 marks)
Complete the indicated part of the function verify_tickers. This function takes in a list of tickers to be
verified, and raises an Exception if any of the tickers provided is not a key of the dictionary returned by
the get_tics function. Further details on when to raise an Exception are provided in the “Notes” section
of the docstring.
An Exception is an action that disrupts the normal flow of a program. This action is often representative
of an error being thrown. Exceptions are ways that we can elegantly recover from errors.
To learn more about raising an Exception, you may refer to the following resource:/ https://www.w3schools.
com/python/gloss_python_raise.asp
Step 7: Complete the verify_cols function (10 marks)
Complete the indicated part of the function verify_cols. This function takes in a list of column names to
be verified, and raises an Exception if any of the column names provided are not found in COLUMNS. Further
details on when to raise an Exception are provided in the “Notes” section of the docstring.
5Step 8: Complete the create_data_dict function (15 marks)
Complete the indicated part of the function create_data_dict. This function is used to transform the data
found in the “.dat” files into a single dictionary.
This function takes in 3 arguments:
- tic_exchange_dic
- A dictionary returned by the get_tics function.
- tickers_lst
- A list containing tickers (as strings) whose data we want to save in the dictionary returned by
create_data_dict.
- col_lst
- A list of the columns (as strings) that we want to save in the dictionary returned by
create_data_dict.
An example of how the returned dictionary should look like when we call
create_data_dict(tic_exchange_dic, ['aapl', 'baba'], ['Date', 'Close']) is provided below:
{
'aapl': {
'exchange': 'nasdaq',
'data': [
{
'Date': '2020-01-01',
'Close': '8.0927',
},
{
'Date': '2020-01-01',
'Close': '8.2784',
},
...
]
},
'baba': {
'exchange': 'nyse',
'data': [
{
'Date': '2017-05-13',
'Close': '3.4939',
},
{
'Date': '2017-05-14',
'Close': '3.5689',
},
...
]
}
}
- Each dictionary found in a ticker’s data list should only contain the columns specified by col_list.
6• The data list for each ticker should contain a dictionary for each line of the “.dat” file for that ticker.
Note: The numbers used in the above example are entirely arbitrary for illustrative purposes only.
Step 9: Complete the create_json function (5 marks)
Complete the indicated part of the function create_json. This function saves a given dictionary into JSON
file. To learn more about how to write data to a JSON file, take a look at the documentation for the following
methods from the built-in Python package json:
- json.dump
- json.dumps
Step 10: Open-ended question (25 marks)
- Explain why to configure as in step 1 (5 marks).
- From step 9, we can obtain a JSON file that contains ticker names, listed exchanges, trading volume,
and stock prices, and compute stock returns from stock prices. Suppose we have an extra variable
to proxy the journalists’ views about the firm performance, and we are interested in evaluating 2
alternative hypotheses:
- Whether the articles journalists wrote are based on:
- Investors’ evaluations of those firms?
- Valuable information beyond firm fundamentals?
- Suppose we have observed the following:
– When the articles contain relatively more negative words,
– Stock returns decrease in the short run with no reversal in the long run.
- Explain which one of the two hypotheses is more likely to be true. Please evaluate both hypotheses.
(15 marks)
- Given the hypothesis you have chosen, evaluate the short-run predictability for the trading volume
with reason(s). (5 marks)
Submit your module
Copy and paste the entire content of your zid_project1.py to Ed and press “Submit”. Ed will not mark
your project automatically or give you any feedback. After you submit, we will be able to mark your project.
7Administrative Guidelines, Additional Hints, Marking
Administrative Guidelines
We will enforce the following:
- This assessment must be completed individually. Failure to complete the assignment on your own may
result in a full loss of marks.
- Late submissions are allowed, but will be penalised according to the rules described in the course
outline.
Hints Your code should be portable, working in a variety of settings. It should be sufficient to copy your code fromPyCharm to Ed for submission. However, as part of this assessment, you need to make sure that your codeworks on our computer as well as on yours.
The following hints should help you correct any portability mistakes:
- The contents of your zid_project1.py module must not contain any direct reference to foldersin your computer. Variables that define paths should not contain any forward or backslashes.Of course, the variables you代 写Combining Data from Multiple Sources defined in your toolkit_config.py module (which you do not have tosubmit) do contain forward or backslashes (depending on your operating system). This is one of thereasons why we created this file to begin with.
- Similarly, you should not include strings with specific tickers in zid_project1.py module (e.g.,“TSLA”, “AAPL”). For instance, you should not create a variable called tickers and then copy thespecific tickers you received in your TICKERS.txt file. Instead, yourcode should read the TICKERS.txtfile, produce a list of tickers, and store that in a variable.
- When writing functions in the file zid_project1.py:
- Do not modify the function names or the parameters.
- Only modify the parts indicated by the “<COMPLETE THIS PART>” tag.
- You do not need to import any other module. Please do not modify the import statements.
- You should not create any additional constants. The constants that exist in the file (ROOTDIR, DATDIR,TICPATH, COLUMNS, and COLWIDTHS) should be edited as instructed.
- The “test” functions are included to help you test the code as you work through the project. Thesefunctions will not be marked, and you may change them as necessary to suit your needs. Test functionsare clearly identified in the file with names starting with _test.
- Use all the parameters in a function declaration. For example, the function get_tics(pth) has the
single parameter pth. Make sure that your function uses this parameter and not a global variable.
- Only submit the zid_project1.py module. Make sure your code works with this module only. No
other modules can be submitted.
How we will mark your assessment
The following parts of this assessment will be marked:
- Location of files and folders (3 marks)
- 8 Set the variables describing the format of the source data (2 marks)
- Complete the get_tics function (15 marks)
- Complete the read_dat function (5 marks)
- Complete the line_to_dict function (10 marks)
- Complete the verify_tickers function (10 marks)
- Complete the verify_cols function (10 marks)
- Complete the create_data_dict function (15 marks)
- Complete the create_json function (5 marks)
- Open-ended question (25 marks)o receive full credit for parts 1 and 2, your variables must:
- Have the correct type (e.g., COLWIDTHS must be a dictionary)
- Include the correct values (e.g., the order of the columns in COLUMNS must match the one specified inyour README.TXT file,
- Follow all the instructions in this file (e.g., no forward or backslashes in TICPATH).To receive full credit in parts 3 - 9, your functions must:
- Return the correct object type (described in the docstring)
- Return the correct information from the TICKERS.txt, README.txt, and “.dat” files you received.
- Not violate any of the rules we specified in this document or in the docstring
- If your function opens a file, you must use a context manager.