r (42739)))/
چهار سال پیش منتشر شده
تعداد بازدید: 428
کد پروژه: 348426
شرح پروژه
#کد_پروژه: 42793
موضوع: پروژه r /
جسارتا من ی پروژه دانشجویی داریم واسه r خواستم ببینم شرایط و هزینه ب چ صورت هستدیتاست هم داریمباید ی مدل رگرسیون باشه
متغیر وابسته و حداقل دو تا explantory والا خیلی عجله ایه اما در کل خیلی پروژه ی خاصی نیست در کل ده صفحه با تفسیر هاش 1. General task
The goal of the assignment is to do a complete regression analysis with all necessary tests and diagnostics. Use the Airbnb_listings.xlsx data set that you can download from Canvas. Conduct all your analyses with R-Studio and combine all output into a pdf document (e.g., using MS. Word). The document needs to feature
The regression analysis needs to be embedded into a short text motivating and describing what you have been doing.
All relevant diagnostics with a brief explanation of what they mean and what their results imply.
A final verdict on whether the results of the regression can be interpreted or not.
2. Detailed description of the task
Define a regression model that you think is interesting to look at on the basis of the available variables. Write a short paragraph on a hypothesis that might motivate the model, i.e., what is the focal explanatory variable and which are control variables helping to isolate the empirical relation of the focal with the dependent variable. The minimum requirements for your model are:
One continuous dependent variable
At least 1 explanatory variable of any kind AND one explanatory variable that is nominal in nature
Select the following variables from the data set: your dependent variable, your (min) 2 explanatory variables, latitude, longitude, and neighbourhood_group_cleansed.
Adapt all variables such that they are numerical in nature. This includes transforming your nominal explanatory variable into a set of dummy variables.
Execute the following code, whereby Your.seed.number is the days-information of your birthday (i.e. birthday on June 27, 1992 implies Your.seed.number=27) and Name.of.your.data is the name of your data set. After the command, your data set will have been reduced to 1000 observations. Do not change this code or the value of your seed number.
set.seed(Your.seed.number)
Name.of.your.data <- Name.of.your.data %>% sample_n(size=1000)
Run your linear regression model and do all necessary diagnostics. In case of a model assumption being violated, try ONE thing to resolve the issue and briefly describe each step. If the “solution” doesn’t solve the issue, pretend in the following that it did. Potential issues of spatial auto-correlation are exempt from this, it is optional to try to solve the issue (see below).
A note on the test of spatial autocorrelation: There is no shape file to extract the neighborhood information from. In fact, your observations are individual flats and houses and not regions. However, for each of these, their location is expressed by means of the geo-coordinates (latitude and longitude in your data). This is very detailed location information from which neighborhood structures can be obtained. Run the following code to create a neighborhood list object (nb).
همیارپروژه, [۱۰/۱۰/۲۰۲۲ ۰۷:۲۱ ق.ظ]
This extracts the two columns of geo-coordinates from your data.
coords <- Name.of.your.data %>% select(longitude,latitude)
This transforms this information into a spatial data object (which is required for the subsequent functions).
coordinates(coords) <- ~ longitude + latitude
It identifies the five observations that are closest to the focal observation, i.e., which five Airbnb objects are closest to the focal. Being geographically the closest is used as the definition of the neighborhood in this case. You can play around with the 5, exploring other sets of neighbors. It is also possible to consider all neighbors that are within a specific geographical distance to an object, i.e., within 100 meters: neig <- dnearneigh(coords, 0, 0.01, longlat = TRUE). However, it is easier to work with a specific number of (nearest) neighbors, as it ensures that each observation has the same number of neighbors.
neig <-knearneigh(coords,5,longlat = T)
Lastly, transforms the k-nearest neighbors set into a neighborhood list that you already know from working with the shape file. neig <- knn2nb(neig)
Note that the order of observations in the neigh object corresponds to that in your data set, so you do not need to do any left_join() merging and can simply use this object in the subsequent operations.
Optional: In case you are encountering a problem of spatial auto-correlation, you can try to create a so-called spatial-lag variable, i.e., a variable that captures some characteristics of the neighboring regions, i.e., that are represented by one of your explanatory variables. The code below allows you to quickly create such a variable using the function lag.listw() of the spdep package. The function requires the specification of the explanatory variable you want to create spatial lag of (one.of.your.explanatory.variables) and the spatial weight object (your.spatial.weight.object) that you’ve created to test for spatial auto.correlation.
spatial.lag.variable=lag.listw(var=one.of.your.explanatory.variables,x=your.spatial.weight.object,zero.policy = T)
3. Assessment and requirements for passing
Pass/fail
A complete document (as pdf) with the corresponding R-code attached. The maximum number of pages (A4 equivalents) is ten including references and visualisations. Overlap with other students' works is limited to parts of the R-code given in the assignment. A plagiarism check will be done and similarities between works will be assessed.
Late submissions will not be considered. اگر زودتر قیمت بدین ممنون میشم این هم فایل دیتاست این خود متن assignment هست
مهارت ها و تخصص های مورد نیاز
مهلت برای انجام
2روز
وضعیت مناقصه
بسته
درباره کارفرما
عضویت سیزده سال پیش
