Items tagged with data-analysis

Feed

hello i was wondering is it possible to find the standard error of the coefficents in the NonlinearFit function within the statistical package? or in any other package for that matter?

does someone have angorithme that can find it? 

i've searched around the internet and it does not seem like maple employs this function at all?

hope someone can help:)

Aggregate statistics are calculated by splitting the rows of a DataFrame by each factor in a given column into subsets and computing summary statistics for each of these subsets.

The following is a short example of how the Aggregate command is used to compute aggregate statistics for a DataFrame with housing data:

To begin, we construct a DataFrame with housing data: The first column has number of bedrooms, the second has the area in square feet, the third has price.

bedrooms := <3, 4, 2, 4, 3, 2, 2, 3, 4, 4, 2, 4, 4, 3, 3>:
area := <1130, 1123, 1049, 1527, 907, 580, 878, 1075,
1040, 1295, 1100, 995, 908, 853, 856>:
price := <114700, 125200, 81600, 127400, 88500, 59500, 96500, 113300,
104400, 136600, 80100, 128000, 115700, 94700, 89400>:
HouseSalesData := DataFrame([bedrooms, area, price], columns = [Bedrooms, Area, Price]);

Note that the Bedrooms column has three distinct levels: 2, 3, and 4.

convert(HouseSalesData[Bedrooms], set);

The following returns the mean of all other columns for each distinct level in the column, Bedrooms:

Aggregate(HouseSalesData, Bedrooms);

Adding the columns option controls which columns are returned.

Aggregate(HouseSalesData, Bedrooms, columns = [Price])

Additionally, the tally option returns a tally for each of the levels.

Aggregate(HouseSalesData, Bedrooms, tally)

The function option allows for the specification of any command that can be applied to a DataSeries. For example, the Statistics:-Median command computes the median for each of the levels of Bedrooms.

Aggregate(HouseSalesData, Bedrooms, function = Statistics:-Median);

By default, Aggregate uses the SplitByColumn command to creates a separate sub-DataFrame for every discrete level in the column given by bycolumn.

with(Statistics);
ByRooms := SplitByColumn(HouseSalesData, Bedrooms);

We can create box plots of the price for subgroups of sales defined by number of bedrooms.

BoxPlot( map( (m)->m[Price], ByRooms), 
deciles=false,
datasetlabels=["2 bdrms", "3 bdrms", "4 bdrms"],
color=["Red", "Purple", "Blue"]);

 

I have recorded a short video that walks through this example here: https://youtu.be/e0pqCMyO3ks

The worksheet for this example can be downloaded here: Aggregate.mw

This is the second of three blog posts about working with data sets in Maple.

In my previous post, I discussed how to use Maple to access a large number of data sets from Quandl, an online data aggregator. In this post, I’ll focus on exploring built-in data sets in Maple.

Data is being generated at an ever increasing rate. New data is generated every minute, adding to an expanding network of online information. Navigating through this information can be daunting. Simply preparing a tabular data set that collects information from several sources is often a difficult and time consuming effort. For example, even though the example in my previous post only required a couple of lines of Maple code to merge 540 different data sets from various sources, the effort to manually search for and select sources for data took significantly more time.

In an attempt to make the process of finding data easier, Maple’s built-in country data set collects information on country-specific variables including financial and economic data, as well as information on country codes, population, area, and more.

The built-in database for Country data can be accessed programmatically by creating a new DataSets Reference:

CountryData := DataSets:-Reference( "builtin", "country" );

This returns a Reference object, which can be further interrogated. There are several commands that are applicable to a DataSets Reference, including the following exports for the Reference object:

exports( CountryData, static );

The list of available countries in this data set is given using the following:

GetElementNames( CountryData );

The available data for each of these countries can be found using:

GetHeaders( CountryData );

There are many different data sets available for country data, 126 different variables to be exact. Similar to Maple’s DataFrame, the columns of information in the built-in data set can be accessed used the labelled name.

For example, the three-letter country codes for each country can be returned using:

CountryData[.., "3 Letter Country Code"];

The three-letter country code for Denmark is:

CountryData["Denmark", "3 Letter Country Code"];

Built-in data can also be queried in a similar manner to DataFrames. For example, to return the countries with a population density less than 3%:

pop_density := CountryData[ .., "Population Density" ]:
pop_density[ `Population Density` < 3 ];

At this time, Maple’s built-in country data collection contains 126 data sets for 185 countries. When I built the example from my first post, I knew exactly the data sets that I wanted to use and I built a script to collect these into a larger data container. Attempting a similar task using Maple’s built-in data left me with the difficult decision of choosing which data sets to use in my next example.

So rather than choose between these available options, I built a user interface that lets you quickly browse through all of Maple’s collection of built-in data.

Using a couple of tricks that I found in the pages for Programmatic Content Generation, I built the interface pictured above. (I’ll give more details on the method that I used to construct the interface in my next post.)

This interface allows you to select from a list of countries, and visualize up to three variables of the country data with a BubblePlot. Using the preassigned defaults, you can select several countries and then visualize how their overall number of internet users has changed along with their gross domestic product. The BubblePlot visualization also adds a third dimension of information by adjusting the bubble size according to the relative population compared with the other selected countries.

Now you may notice that the list of available data sets is longer than the list of available options in each of the selection boxes. In order to be able to generate BubblePlot animations, I made an arbitrary choice to filter out any of the built-in data sets that were not of type TimeSeries. This is something that could easily be changed in the code. The choice of a BubblePlot could also be updated to be any other type of Statistical visualization with some additional modifications.

You can download a copy of this application here: VisualizingCountryDataSets.mw

You can also interact with it via the MapleCloud: http://maplecloud.maplesoft.com/application.jsp?appId=5743882790764544

I’ll be following up this post with an in-depth post on how I authored the country selector interface using programmatic content generation.

 

how to call matlab to run k means on the data and

how to show which circle do the row of data belong to?

hi,

 

I am trying to implement some data-intensive algorithms like: comuting similarity scores, clustering etc. How efficient Maple is in doing these operations? I mean is it comparable to MATLAB in these operations? What toolboxes are useful in these? I am trying to build a recommender system.

 

I am using Maple18 and not sure if Maple has any advantage in implementing these type of algorithms over MATALB??

 

thanks

siba

Hello, i am working on creating a smooth curve from my data points: [0.2e-1, .158], [0.2e-1, .159], [0.3e-1, .161], [0.3e-1, .164], [0.3e-1, .166], [0.4e-1, .169], [0.6e-1, .173], [0.8e-1, .178], [.1, .185], [.11, .187], [.14, .193], [.19, .2], [.28, .21], [.38, .223], [.44, .233], [.58, .244], [.82, .256], [1.4, .278], [1.71, .281], [1.78, .282], [1.78, .282], [1.81, .282]

however, I am not sure how to to create the smooth curve formula for said data points. I tried using polycurve, but it didn't spit out any equations.

Page 1 of 1