数据分配作业代写 DATA ASSIGNMENT 英国代写

DATA ASSIGNMENT

数据分配作业代写 This section uses the prefectural level monthly dataset created in Problem 1. We are interested in documenting the variation

OVERVIEW  数据分配作业代写 

This assignment is inspired by a recent project of EPIC-China’s faculty. For this task, you will be asked to conduct data cleaning and preliminary analysis to investigate the temperature-health
relationship in Japan. You will manage weather variables collected from nearly 800 monitoring stations, and prefectural level mortality data from 2001 to 2015.

When you are done with Problem 1-3, please send in the following items.

• Ÿ Well-commented code files in the language of your choice (.do files for Stata, .R or .Rmd for R, .py or Jupyter Notebook for Python, etc.)


• Ÿ Final dataset produced from Problem 1


• Ÿ Final graphs and tables produced from Problems 2 and 4


• Ÿ A short document answering the questions raised in Problem 3

This task should take 4 hours or less, but you will have 12 hours in total to complete it. The goal of the test is to give you an opportunity to demonstrate your coding competency and your conceptual understanding of empirical economics research. A perfect grade is not a prerequisite for consideration for the position. If you find any of these instructions to be confusing, please proceed in a way that you find relevant and reasonable, and list your assumptions in your writeup. Please do not email us for any clarification questions, as we will likely not be able to respond in a
timely fashion.

PROBLEM 1. CLEANING THE DATA  数据分配作业代写

The central task of this section is to collapse the station-level weather dataset into a prefecturallevel dataset (Japan has 47 prefectures). The algorithm to match station-level data to a prefecture
is as follows: take a weighted average of monthly mean temperature and monthly total precipitation, for all monitoring stations within each prefecture. The weights are the inverse of the squared
distance from the prefectural population center.

The dataset “prefecture” includes a geo-coded population center in each prefecture. The “temperature” and “precipitation” datasets include monthly mean temperature and monthly total precipitation measured at each monitoring station. To create the weighted average, you will have to take distance from the prefectural population center to each monitor station. For Stata users, the geodist and vincenty commands could be useful.

The final product from this section is a prefecture-level dataset from 2001-2015 with information on temperature and precipitation. Please include both the constructed dataset and
your codes in your final submission.

PROBLEM 2. DATA EXPLORATION

This section uses the prefectural level monthly dataset created in Problem 1. We are interested in documenting the variation in temperature and precipitation.

1. You will use five prefectures (1.Hokkaido, 13.Tokyo, 27.Osaka, 40.Fukuoka, and 47.Okinawa). For each prefecture, create a time-series dataset of monthly temperature and precipitation averaged over 15 years. The dataset for Hokkaido should have 2 variables, and each variable should have 12 observations corresponding to each month.


2. Using this time series data, create two scatterplots depicting temperature and precipitation distribution by month for Tokyo prefecture.


3. Format your graphs so that they are of publication quality and save them as .pdf files.


4. Create one publication-quality table of summary statistics for five prefectures. Please show the average temperature and precipitation in four periods: 2001-2015, 2001-2005, 2006-
   2010, and 2011-2015. In the table, please show the data on five prefectures * four periods * two variables in a stylish way.

PROBLEM 3. DESIGNING MODELS FOR CAUSAL INFERENCE  数据分配作业代写

You are interested in using this dataset to explore the impact of temperature on prefecture-level mortality in Japan. Note that mortality data is yearly-level.

1.  Write down a simple OLS model regressing mortality on temperature.


2.  The effect of temperature on mortality is very likely nonlinear. How would you modify your model above to estimate any potential nonlinearities?


3. Write down the condition for the estimators to be unbiased. Explain the main identification threats to obtain unbiased estimators.


4. How could you use a) fixed effects and b) other control variables to help address the  identification threats?


5. Write down a full econometric model addressing the concerns raised in questions 2 to 4.

PROBLEM 4. ESTIMATION AND INTERPRETATION

Estimate the models (1 and 5) in Problem 3 and create one publication-quality table reporting the regression results. Interpret the estimated coefficients and standard errors within a page.
For this task, use the “Mortality_rate” dataset that covers the yearly prefectural-level mortality rate per 1,000 and merge it with the weather data.

商科代写 cs代写 法律学代写 经济学代考_经济学作业代写 艺术代写 心理学代写  哲学代写 伦理学代写 体育学代写 化学代写 教育学代写 医学代写 历史代写