Finder Maker

GeoCommons Maker User Manual

What is GeoCommons Maker?

Maker! allows you to create professional-looking interactive maps from a wide variety of geographic data in only a few minutes. Maker! operates right in your web browser and allows you to easily publish and share maps as embed images, URL links, or PDFs. You don’t need to be an expert in cartography to use Maker!. It will walk you through the map-making process, highlighting some important decisions and providing suggestions. In no time, you’ll be making great looking maps quickly and easily.

Because Maker! is fully integrated with Finder!, it’s very easy to find and add different data to your map. You can also add your own data to these maps by uploading it through Finder!. Maker! allows you to combine different data together to map multiple themes at the same time. For example, you could make a map to show how population density and home prices are related. It is these kinds of multivariate thematic maps that really allow you to explore data and uncover new insights.

Navigating Maker! is similar to navigating Finder!. Maker! is divided into the following sections: Maker! Homepage, My Maps, Profile, and Make a Map. Each of these sections is explained below.

Maker! Homepage contains the Find a Map and Browse sections, which allow you to search for previously created maps from other users by keywords or categories. The homepage also displays a count of your maps, a link to "My Maps," a video introduction to Maker!, and an Explore Geodata button, which links back to Finder!.

My Maps is a collection of maps you have created. From here you can view, edit, share, and delete your maps.

Profile is where you can find information about your GeoCommons account and edit your account information.

Make a Map allows you to build a new map.

Making Maps with Maker!

Broadly speaking, the map-making process in Maker! occurs in two steps. First, you decide what kind of map you want (of the 4 listed below); second, you decide how you want that map to be styled. Maker! lets you customize the look of the map, such as the colors, line styles, types of symbols, number of data classes, and so on. If you’re in a hurry, Maker! has some smart default choices so you don’t have to sweat the details. We encourage you to explore your options to get the most out of your map and let it tell the story you want it to.

Kinds of Maps

With Maker! you can create reference maps, visual theme maps, and data analysis maps.

Reference maps show the location of geographic features, e.g. plotting the paths of the major waterways of North America or the location of oil wells in a country. In this case, you’re using simple geometric symbols like dots, pushpins, or dashed lines to show where something is located. You should use this map when you don’t wish to associate values with your data points, i.e. just school locations, not the student populations at each location.

Visual theme maps use shapes or symbols to display values associated with locations. In Maker! you can choose to display the data through colors or shapes:

Color maps (also known as choropleth maps): If your data is attached to areas, such as states or countries, you can show the data using colors to represent different values, e.g. visualizing per capita income for the countries of the world using a light-to-dark color sequence. Note: If your data is attached to points or lines, color maps are not an option.

Size maps (also known as proportional or graduated symbol maps): If your data is attached to areas or points, you can use the size of symbols to represent different values, e.g. visualizing attendance numbers for sports teams or concert venues.

When selecting which visual theme map to create, keep in mind:

1. Because size maps use the relative size of the shape to communicate, they can become quite dense and often harder to read if you have many large, overlapping symbols.

2. One advantage of size maps is that areas with the largest numbers really jump out at you, whereas these places might be missed on a color map if the size of the country/state/etc. is small.

Data Analysis maps are advanced versions of visual theme maps. They allow the user to perform the following data analysis tasks:

Mapping the difference between two variables: for instance what is the difference in population for all counties in 2008 and 1980

Mapping the correlation between two variables: for instance what is the correlation between population in all counties in 2008 and the number of immigrants in all counties in 2008.

Searching for Data to Map

Since Maker! is fully integrated with Finder!, it’s easy to find and add different data sets to your map without ever leaving the Maker! application. If you are starting from the Maker! homepage, click Make a Map to enter the main Maker! interface. Then click on the Add Data button, which will bring up the Finder! Express portal. From here you are able to access all of the data in Finder! You can search for specific keywords or phrases, or you can browse by category.

Alternatively, you can click on the Finder! button in upper left corner, which will take you to the Finder! homepage. From there, you can search for data and once you find it, click Make a Map to return to Maker!

Searching for Existing Maps

If you’d rather skip the map making process and search for existing maps, Maker! allows you to search and browse within your own appliance or across multiple GeoCommons appliances through federated search.

The search function is as easy as typing a keyword or phrase into the 'Find a Map' bar on the Maker! homepage. The results page will display any datasets that match your keywords in either their title and/or tags.

The browser function is as simple as clicking on a subcategory in the browser section of the Maker! homepage. Click the category link to see the category menu and then select either the main category or any of the subcategories.

Sort By

This feature allows you to arrange your search results by relevance, age and name. Your initial search results will automatically be sorted based upon relevance to the search term or terms. Maker! will display the results that it believes best fit your search criteria. Your results can also be sorted by date of upload. Simply, click the Newest First button to sort your search results to display the most recently uploaded maps. Select the Oldest First button to show the opposite. Your results can also be placed in alphabetical order by selecting the Name – A to Z or the Name – Z to A buttons. This will place your results in alphabetical order based upon the title of the map layer.

How do I start? The Brewer Module

Once you’ve found the data you want to map, you’ll see a series of ‘Map Brewer’ screens that walk you through the initial choices in the map-making process. The content on these screens varies based on what kinds of maps are available for the data you want to map. Be sure to click on the “learn more” links to get an explanation of what to do at each step. You can always back up to a previous screen and change your selections. As you make choices in the Brewer, the map will automatically preview these choices for you. Since designing a map is a highly visual process, rely on these previews to help create the look you want.

Map Brewer Screen 1: Select the type of map you wish to create. If you select reference map, the Brewer ends and you will be taken directly to the map to apply your own styles. If you are making a thematic map, the brewer will take you through screens 2-5.

Map Brewer Screen 2: Select the attribute(s) you would like to display on your map. If you’ve chosen to make a data analysis map, you’ll have to chose between differencing and correlation.

Map Brewer Screen 3: If you are mapping data attached to area units such as counties or states, Brewer 3 will ask you if you would like to display your data as colors or sizes. If your data is only attached to point locations, you will automatically skip this step.

Map Brewer Screen 4: Select the classification method. Included are four of the most common data-classification methods used in cartography. If you are making a proportional symbol map, you have the option to use no classification (on an unclassed map, the symbols will be drawn directly proportional to their data value). If you are unsure which one to use, try matching the shape of the distribution of your data above to one of the four choices below. If you don’t see a perfect fit you can always change this later (you can even do your own manual classification from the Layers palette).

Map Brewer Screen 5: Style your data.

Working in the Main Maker! Interface

In the main Maker! interface, you get to further explore your data, add more data, and customize your map through the main map control buttons (Add Data, Layers, Basemap, Details, Link), and other features.

The Add Data button brings up the Finder! Express portal. From here you are able to access all of the data in Finder! You can search for specific keywords or phrases, or you can browse by category.

Note: If the Add Data button is grayed out and cannot be selected, that indicates that you have reached the data limit on your map.

The Layers Palette is the heart of the Maker! mapping system and allows you to make a number of important cartographic decisions. You can 1) set the number of classes on your map, 2) pick a data classification method, 3) manually choose your own data classes, 4) see the relative number of observations in each class, and 5) standardize your data by an additional variable.

Note on data standardization: If you are mapping data on a visual theme map, we strongly encourage you to consider standardizing your data. With a visual theme map you should not map raw counts or totals (e.g., # of people) but rather rates (e.g., # of people per square mile). This allows you to see the relative differences among places and compare small places and big places more accurately (e.g., California and Rhode Island). It’s not always possible to do this, but consider standardizing your data based on the population of each region (x per person) or by its area (x per sq. mile).

The Basemap Palette is where you select the background tiles for your map. You can choose from road, hybrid, satellite, terrain, and solid.

The Styles Palette allows you to customize the symbols used on your map, including the size of the symbols, their color and transparency, the shape and style of the icons and lines, and the color sequences used on choropleth maps. Open the Styles Palette by clicking on the color/sizes icon in the Layers Palette. The style palettes have been carefully designed to work well across a variety of background images (satellite, roads, terrain). We encourage you to explore these options to create a unique look for your data.

The Map Details button brings up a window where you can add some details/background information about your map. You can write a short description telling the story behind the map, and the dataset(s) used to create the map. Tags can also be added, which will allow other users to easily find your map.

Be sure to select the Done button after making any changes in order to save them.

The View the Map link in the bottom right hand corner takes you to the Published view of the map (as opposed to the Editing view).

After clicking the View the Map link, users have the options to:

3D mode requires the installation of a Google Earth plug in. Maker! will provide a link to download the proper plug in (if it has not been previously installed on your system).

By selecting the Get Google Earth Plugin now link, download and installation will begin automatically. After installation users can view their data in an interactive 3D environment.

Other features of the main interface include "clicky" data and an expandable legend.

Embedding a Map

Maps built in Maker can be easily embedded into other websites or collaboration sites. To embed a map, click the "Details" button when viewing the map you want to embed. In the now visible "Details" window, you will see a link that says "Want to embed this map into your site?". Clicking on this link will reveal a text box that has some HTML code.

Highlight and copy this embed code, either by right-clicking with your mouse, or pressing <Control>-C on your keyboard. Then you can open up your web page editor or content management system and paste this embed code into your site, again using the right-click on your mouse or <Control>-V.

The Maker map is by default 400 pixels high. You can modify this default height and width and changing the numbers in the <style> section of the embed code. You can use percentage, or fixed pixel values. For example, to set the height to 300 pixels and the width to 500 pixels - the styling would look like:

height: 300px; width: 500px

Advanced Features

Data Analysis

Data analysis allows users to perform two functions: 2) the difference between two attributes; 3) correlation between two attributes.

Map the Difference Between Two Attributes

Mapping the difference between two attributes – is a simple calculation that shows the difference between two values within a dataset. The user will choose both attributes they wish to analyze; the first being the attribute they wish to subtract from the second attribute. Maker! will then calculate the difference between the selected values. The results will then be paired to a visual representation that covers the range of resulting values. The final visualization will show the results of the difference analysis.

It is important to note that the calculation is performed on two different attributes for the same location.

Using the Analysis Function

The analysis function is designed to be simple and intuitive. It is built directly into the Maker! Brewer process for creating maps. This walk through will show the differencing process from Finder! through Maker!.

Mapping the difference between two attributes is particularly useful when a user wants to see the change from one time period to another (Temporal Analysis) – for instance, the difference between unemployment by state from 2008 and 2009. The resulting map would show states that have had increases in unemployment and those that have had decreases in unemployment. This provides a concise view of change in the data between the two time periods.

This walkthrough will map the difference between Unemployment numbers from September 2008 to September of 2009, and work through the entire Finder! and Maker! workflows.
Starting on the Finder! homepage search for a dataset pertaining to Civilian Labor Force figures by state.

Type the keywords into the search bar and click the Search button. The following screen will display a list of results that match the search terms. Select the dataset to be mapped. In this case:

Civilian labor force and unemployment by State, United States, 10/28/2009

Clicking on the Make a Map button will export the dataset directly into Maker! and begin the map Brewer process. The Map Brewer will walk users through the map creation process.
On the first Brewer screen, select the Data Analysis option and click the Next button.

On the second Brewer page, choose the Map the DIFFERENCE BETWEEN 2 attributes option. The following two boxes represent the two attributes that will be included in the calculation. Choose the attribute that is to be subtracted in the first box, and the attribute that the first is to be subtracted from in the second box.

This is the most important step of the process, as this step is where users tell Maker! exactly what attributes to use in its calculations. Be sure to select the appropriate attributes to achieve the desired results.

After selecting the two attributes to be included in the analysis, click the Next button. The third Brewer page allows users to select between two types of visualizations: Colors and Sizes.

Please note that in some cases only one option will be available. This is dependent on the type of data file the map is being created from.

After selecting the proper visualization, click the Next button. The fourth Brewer screen allows users to choose how the data is classified. Select between Quantile / Equal Interval / Standard Deviation / and Maximum Breaks.

The fifth and final Brewer screen gives users the option to customize the visualization of their data. Select a color ramp, and click the Finish button.

After completing the Brewer process, please finish the map creation process by titling / describing / and saving the map (please see the Maker! User Manual for further instruction).

Map the Correlation Between Two Attributes

Correlation Analysis allows a user to see the relationship between two variables, and how well one explains the other – called a Pearson’s correlation. This type of correlation indicates the strength and direction of a linear relationship between the two variables.

To perform the correlation analysis the user must select a dependent and an independent variable within the dataset.
• Dependent: what you are trying to predict.
• Independent: the constant being used to predict with.

After selecting the appropriate attributes, Maker! will perform the analysis and display the results for each geographic location. The results will be paired to a color ramp, which will visually indicate the strength of the correlation in each geographic area.

Maker! also plots the correlation in a scatter plot found in the Layers bar. Those features that are highly correlated will be closest to the diagonal axis line in the scatter plot and those least correlated (outliers) will be furthest from the diagonal axis.
• Zero Residual: (Close to the Axis): Expected given the overall correlation
• Positive Residual: (Above the Diagonal Axis): Dependent variable is greater than predicted
• Negative Residual: (Below the Diagonal Axis): Dependent variable is less than predicted

Using the Analysis Function

The analysis function is designed to be simple and intuitive. It is built directly into the Maker! Brewer process for creating maps. This walk through will show the process behind mapping the correlation between two layers within Maker!

This example will predict the correlation between the Hispanic population and the total population by census tract in New Hampshire. Figures are from the 2000 census.
To begin, find a data layer to map the correlation between two layers. This can be done by searching through Finder! and or Maker!. After finding an appropriate data layer, export it into Maker! by choosing the Make a Map option associated with the layer. This will begin the Map Brewer process, which walks users through a step by step map creation process.
Data layer used in this walk through: US Census, New Hampshire by Tracts, Demographics 2000

Select the Data Analysis option on the first Map Brewer Screen, then click the Next button. The next page will be the second Map Brewer screen, where users can choose to:

• Map the DIFFERENCE BETWEEN 2 attributes
• Map the Correlation BETWEEN 2 attributes

Choose the: Map the DIFFERENCE BETWEEN 2 attributes option. The following two boxes represent the two attributes that will be included in the calculation. Choose the attribute that is to be the Independent value in the first box, and the attribute that is to be the dependent in the second box.

• Independent: Population
• Dependent: Hispanic

This is the most important step of the process, as this step is where users tell Maker! exactly what attributes to use in its calculations. Be sure to select the appropriate attributes to achieve the desired results.

After selecting the two attributes to be included in the analysis, click the Next button. The third Brewer page allows users to select between two types of visualizations: Colors and Sizes.

Please note that in some cases only one option will be available. This is dependent on the type of data file the map is being created from.

After selecting the proper visualization, click the Next button. The fourth Brewer screen allows users to choose how the data is classified. Select between Quantile / Equal Interval / Standard Deviation / and Maximum Breaks.

The fifth and final Brewer screen gives users the option to customize the visualization of their data. Select a color ramp, and click the Finish button.

After completing the Brewer process, please finish the map creation process by titling / describing / and saving the map.

Analyzing the Results

In addition to the correlation number the residuals for each feature will be drawn on a scatter plot.

Scatter plots are a simple visualization of the relationship between two variables showing data points on a two dimensional graph. The Dependent variable is plotted on the y axis, while the Independent variable is plotted on the x axis. The placement of each data point within the scatter plot provides the following information about the relationship between the two variables:

• Strength
• Shape – Linear / Curved / etc.
• Direction – Positive / Negative
• The presence of outliers
• Correlation between the two variables (clustering of points)

Those features that are highly correlated will be closest to the diagonal axis line in the scatter plot and those least correlated (outliers) will be furthest from the diagonal axis. If you click on any of the points in the scatter plot they will be highlighted on the map in a turquoise color. In addition the features on the map will be colored to indicate whether they are closely correlated or not.

The Analysis Function

The analysis feature allows users to delve even deeper into their datasets and maps. Users can now visually represent the interactions between multiple attributes within their datasets.

Advanced Analysis

Map the correlation between two attributes – A Pearson’s correlation indicates the strength and direction of a linear relationship between two random variables. It is very widely used in the sciences as a measure of the strength of linear dependence between two variables, giving a value somewhere between +1 and -1 inclusive. The coefficient ranges from −1 to 1. A value of 1 shows that a linear equation describes the relationship perfectly and positively, with all data points lying on the same line and with Y increasing with X. A score of −1 shows that all data points lie on a single line but that Y increases as X decreases. A value of 0 shows that a linear model is not needed – that there is no linear relationship between the variables.

In addition to calculating the correlation the analysis will also map the residuals for the relationship between each feature in the correlation. For instance if the user was running a correlation between the population never married and the total population for each state in the USA the correlation would be calculated and then the residual for each individual state
would be mapped – illustrating how strong the correlation was between the two attributes for each state.

The residual associated with a state indicates how close the correlation of the two attributes are to the correlation that is quantified for all states. A residual close to zero that the correlation is what we would expect to see given the overall correlation across all states. A positive residual (above the diagonal axis) means that the attribute for the dependent variable is greater than what would be predicted from the correlation coefficient. A negative residual is just the reverse where the dependent variable is less than what would be predicted from the correlation coefficient.

In order to perform this function the user must choose an independent and dependent attribute. The dependent variable is what you are trying to predict and the independent variable is the constant you are using to predict with. So, in the example above we would be trying to predict the never married population using total population. Never married population would be the dependent attribute and total population would be the independent attribute. Once you have chosen you independent and dependent attributes the analysis will be run giving you a
correlation number (between one and negative one).

The more detailed mathematical explanation for the analysis illustrated in the correlation function follows below:

The Pearson’s Correlation is a single metric that measures correlation. The residuals mapped in the tool on the other hand come from a regression. In Particular, they are the residuals associated with a standardized regression model. If a non standardized model is as follows:

Y=alpha+bet*X
The standardized model is:
Y=bet(std)*X, (note: in this model you have suppressed the intercept term)

Where:
Bet(std) = beta*(std deviation of y divided by the standard deviation of x).
Bet(std) is equivalent to a Pearson’s coefficient.
Now, to get the residuals you simply subtract the Expected value of Y from the actual value of Y. The expected value of Y is what you would get if you plugged the actual X into the estimated standardized regression.

Frequently Asked Questions (FAQ)