You are on page 1of 34

GISC9308 Deliverable 2

Introduction to Spatial Analyst

Photography Provided by: Lisa Atkinson

Lisa Atkinson Niagara College 2/22/2013

February 22, 2013 GISC9308-D2 Ian D. Smith M.Sc., OLS, OLIP, EP GIS-Geospatial Management Niagara College 135 Taylor Road Niagara on the Lake, ON L0S 1J0

Dear Mr. Smith, RE: GISC9308 Deliverable 2 Introduction to Spatial Analyst Please accept this letter as my formal submission of Deliverable 2: Introduction to Spatial Analyst for GISC 9308 Spatial Analysis Statistics. This deliverable contains one document demonstrating thorough comprehension of spatial analysis techniques, specifically, the appropriate utilization of weighted multi-criteria evaluation, and fuzzy logic. The investigation presents a logical flow of data manipulation, resulting in carefully crafted map layouts and model building. The ideal placement of fern reintroduction is limited to subsections of the Niagara Escarpment, and the sloping landscape located alongside the railway, in proximity to the Niagara on the Lake, Niagara College campus. Furthermore, the preferred method of suitability analysis is the fuzzy analysis technique. Should you have any questions or concerns regarding the enclosed document, please contact me at your convenience by email at lisaclaire87@gmail.com or by phone at (705) 499-6768. Thank you for your time and attention. I look forward to your comments and suggestions. Sincerely,

Lisa Atkinson BA (Honours) Geography, Nipissing University Geomatics Certificate GIS-GM Certificate Candidate
L.A./l.a.

Enclosures:

1) Introduction to Spatial Analyst

GISC9308-D2 February 22, 2013 Lisa Atkinson Executive Summary The majority of spatial decisions is not influenced by a single criterion, but must account for a collection of variables and conditions. As spatial investigations become more complex, platforms such as ArcGIS, must adapt and accommodate the intricate demands of the GIS specialist. The weighted multi-criteria evaluation analysis, considers each criterion, or variable, as a percentile of importance, in regards to the other criteria. However, reality is most often defined in terms of partial memberships, or varying degrees of truth. For this reason, fuzzy membership overlay analysis is becoming increasingly adopted, as the desired data manipulation tool. Both techniques are investigated, and compared, in an attempt to identify the most suitable locations, in proximity to Niagara College Glendale Campus, to reintroduce the rare (and completely fictitious), Finalys Fabulous Fictitious Fern. The ideal placement of fern reintroduction is limited to subsections of the Niagara Escarpment, and the sloping landscape located alongside the railway, in proximity to the Niagara on the Lake, Niagara College campus. Furthermore, the preferred method of suitability analysis is the fuzzy analysis technique.

i|Page Introduction to Spatial Analyst

Table of Contents

GISC9308-D2 February 22, 2013 Lisa Atkinson

Executive Summary ....................................................................................................................................... i 1.0 Introduction ............................................................................................................................................ 1 2.0 Background ............................................................................................................................................. 1 3.0 Goal Statement....................................................................................................................................... 3 4.0 Methodology .......................................................................................................................................... 3 5.0 Weighted Multicriteria Evaluation ........................................................................................................ 8 5.1 Evaluation Logic .................................................................................................................................. 8 5.2 Data Transformations ....................................................................................................................... 10 5.3 Results ............................................................................................................................................... 15 6.0 Fuzzy Overlay Analysis ......................................................................................................................... 17 6.1 Evaluation Logic ................................................................................................................................ 17 6.2 Data transformations ........................................................................................................................ 19 6.3 Results ............................................................................................................................................... 19 7.0 Recommendations ............................................................................................................................... 21 8.0 Conclusions ........................................................................................................................................... 25 9.0 Bibliography.......................................................................................................................................... 26 Appendix A .................................................................................................................. Data Model Diagram Appendix B........................................................................................................ Original Terms of Reference

List of Figures

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 1: Study Area ..................................................................................................................................... 2 Figure 2: Data Correction ............................................................................................................................. 4 Figure 3: TIN to Raster.................................................................................................................................. 6 Figure 4: Reclassification of Hillshade ....................................................................................................... 11 Figure 5: Reclassification of Aspect .......................................................................................................... 12 Figure 6: Reclassification of Slope ............................................................................................................. 13 Figure 7: Soil Type Reclassification ............................................................................................................ 14 Figure 8: MCE Results ................................................................................................................................. 16 Figure 9: Membership Types ...................................................................................................................... 18 Figure 10: Fuzzy Membership Data Transformations .............................................................................. 19 Figure 11: Fuzzy Analysis Results ............................................................................................................... 20 Figure 12: Recommended Fern Planting Sites ........................................................................................... 22 Figure 13: Example of Fuzzy Overlay Analysis ........................................................................................... 24

List of Tables
Table 1: Ranking Convention ....................................................................................................................... 8 Table 2: Application of Ranking Convention ............................................................................................... 9 Table 3: Fuzzy Membership Parameters ................................................................................................... 17 Table 4: Suitability Area Comparison ........................................................................................................ 23

GISC9308-D2 February 22, 2013 Lisa Atkinson 1.0 Introduction The majority of spatial decisions is not influenced by a single criterion, but must account for a collection of variables and conditions. As spatial investigations become more complex, platforms such as ArcGIS, must adapt and accommodate the intricate demands of the GIS specialist. Spatial Analyst, in ArcGIS, allows for the development, and execution, of weighted multi-criteria analysis. However, in practical terms, real world situations do not adhere to computer calculated probabilities. Fuzzy membership and overlay analysis tools, in ArcGIS, provide more flexible weighting of evidence and criterion combinations (Raines, 2012). The most significant economic implication of utilizing precise spatial analysis techniques is a decrease of project cost. Hence the growing demands for the application of fuzzy logic to create descriptive, although imprecise, spatial solutions (Raines, 2012).

2.0 Background In order to identify potential areas to reintroduce Finalys Fabulous Fictitious Fern, within the natural environment surrounding the Niagara College Glendale Campus, the following data is provided by Niagara College, to perform multi-criteria evaluations. The provided elevation markers and break lines are utilized to create a TIN, from which aspect, hillshade and slope, are modeled. GC_Campus_North_Mass_Points.shp GC_Campus_South_Mass_Points.shp GC_Campus_North_Breaklines.shp GC_Campus_South_Breaklines.shp

Additionally, soil data, in a polygon shape file, denoted as GC_Area_Soils.shp, is available for the study area extent. Esri imagery is utilized as a base map, to provide spatial reference and understanding of the study area. All data is projected to a common display using UTM NAD 83 Zone 17N. The study area is defined by Figure 1, shown on the following page:

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 1: Study Area

GISC9308-D2 February 22, 2013 Lisa Atkinson 3.0 Goal Statement This investigation facilitates the application of complex analysis, and problem solving techniques, in order to complete multi-criteria evaluations. Two approaches are executed to identify the most suitable locations, in proximity to Niagara College Glendale Campus, to reintroduce a rare (and completely fictitious) species of fern: weighted multi-criteria evaluation, and fuzzy overlay analysis.

4.0 Methodology As an application of GIS best practices, the original data, provided by Niagara College, is examined in order to identify missing data, incompatible projections, and accurate spatial placement. All provided data is complete and accurate with the exception of the soil polygon shapefile. Additionally, the soil type data requires a transformation of projection, from UTM NAD 27 Zone 17N, to the common projection utilized by the remaining data files: UTM NAD 83 Zone 17N. Figure 2 displays the incorrect placement of the soil polygon shapefile, in reference to the Esri imagery basemap, and the corresponding transformation, completed via Editor Toolbar, in ArcGIS.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 2: Data Correction

GISC9308-D2 February 22, 2013 Lisa Atkinson

Unfortunately, Finlays Fabulous Fictitious Fern is no longer abundant. The College campus is now an ecologically stable environment, and productive for fern growth. Utilizing expert knowledge, several criteria are identified, and compiled to perform suitability analyses. The criteria are outlined within the Original Terms of Reference, located in Appendix B of this document.

A TIN is created, via 3D Analyst - Create TIN tool. A TIN, or Triangular Irregular Network, is a digital means to represent surface morphology, by connecting point data containing elevation attributes (Esri, 2012). The elevation marker point data, provided by Niagara College, is utilized to create a TIN, which is transformed to a raster, using 3D Analyst TIN to Raster tool, depicting elevation in meters. The creation of a raster elevation model is necessary to render hillshade, slope and aspect models, three criteria variables defined by the Original Terms of Reference, located in Appendix B. Figure 3 displays the appropriate transformation of TIN to elevation raster model, on the following page.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 3: TIN to Raster

GISC9308-D2 February 22, 2013 Lisa Atkinson

This methodology outlines the necessary data preparation, required prior to site suitability spatial analysis. A full account of methodology, and data transformation, are defined by the two data model diagrams, located in Appendix A: Data Model Diagram, of this document. The remainder of this report outlines two spatial analysis techniques, and provides fern reintroduction recommendations.

GISC9308-D2 February 22, 2013 Lisa Atkinson

5.0 Weighted Multi-Criteria Evaluation A weighted multi-criteria analysis evaluates a compilation of variables, in order to acquire spatial suitability (European Commission, 2007). However, variables may vary, in terms of importance, within the decision making process. Therefore, each variable is assigned a weighted average, based on expert knowledge, in order to create hierarchical solutions (European Commission, 2007). 5.1 Evaluation Logic The defined criteria for optimal fern growth conditions, in regards to hillshade, aspect and slope, outlined within the Original Terms of Reference, located in Appendix B, are simply a compilation of vague descriptions. This is less than desirable for the execution of a weighted multi-criteria evaluation. Logically, dissimilar values cannot be evaluated, or coalesced. Therefore, a ranking convention is utilized to create common suitability descriptions. This allows for accurate evaluations, as values defined by a common scale are compared. Table 1 outlines the ranking convention, as applied to all criteria, in order to transform vague description, to conformal values.

Table 1: Ranking Convention

Rank Value Description No Data No Data 0 Unsuitable Location 1 Sparse Growth 2 Moderate Growth 3 Optimal to Thrive

Table 2, located on the following page, displays the application of the ranking convention to all variables defined within the Original Terms of Reference, located in Appendix B.

GISC9308-D2 February 22, 2013 Lisa Atkinson


Table 2: Application of Ranking Convention

Data Value Criterion 0-180 Hillshade 180-195 195-220 220-235 235-255 0-247 degrees Aspect 247-292.5 degrees 292.5-337.5 degrees 337.5-360 degrees 0-20 degrees Slope 20-35 degrees 35-45 degrees 45-55 degrees 55-65 degrees 65-90 degrees Escarpment Soil Type Toledo/Cashel Beverley/Peel Haldimand/Lincoln Urbanized/Not Mapped

Description Will not grow Will grow but not thrive Will thrive Will grow but not thrive Will not grow No growth Very sparse growth Heavy growth Very sparse growth No growth Sparse growth Moderate growth Heavy growth Sparse growth No growth Best suited Reasonably well suited Very poorly suited Not at all suited Not suitable

Conformal Rank 0 1 3 1 0 0 1 3 1 0 1 2 3 1 0 3 2 1 0 No Data

GISC9308-D2 February 22, 2013 Lisa Atkinson 5.2 Data Transformations In order to produce a weighted multi-criteria evaluation, the elevation raster is utilized, via Spatial Analyst - Surface tool, to create a hillshade, aspect, and slope model. Each model is subject to analysis, to determine suitable sites for fern reintroduction. The defined ranking convention is applied to each model, via Spatial Analyst Reclassification tool, resulting in criteria site suitability images. Hillshade is a 3D model of the earths surface, depicting a shaded image based on the suns position in the sky (Esri, 2012). The hillshade is modeled based on the absolute minimal amount of shadow possible. This constraint ensures that site suitability, based on the hillshade criteria, will never deplete in correspondence to altering seasons. The hillshade parameters are defined as follows: Date: June 21 (Summer Solstice, and highest possible sun position relative to horizon); Altitude: 65.98 degrees; Azimuth: 177.1 (NOAA, 2012) The parameters for aspect and slope are defined by the Original Terms of reference, located in Appendix B, of this document. In order for the soil type data to conform to the ranking convention, a new attribute table field is created, to store rank values. The soil type polygon shapefile is reclassified, based on the newly created field. Each of the defined criteria contributes to the growth and successful reintroduction of Finlays Fabulous Fictitious Fern. Figure 4 to 7, outline the reclassification results, and the logical acceptance of these results using feature comparison.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 4: Reclassification of Hillshade

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 5: Reclassification of Aspect

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 6: Reclassification of Slope

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 7: Soil Type Reclassification

GISC9308-D2 February 22, 2013 Lisa Atkinson 5.3 Results Based on expert knowledge, the criteria are ranked by the following statement: The slope criterion is twice as important as slope direction, while slope is three times as important as soil type; soil type and hillshade share the same importance. The qualitative statement is solved to generate the following weighted percentiles: Slope: 46% Aspect: 23% Soil Type: 15.5% Hillshade: 15.5% Hence, the following formula is utilized within Spatial Analyst Raster Calculator. The formula accounts for the weighted percentiles defining each criterion, as well as the elimination of unsuitable areas.

((

( ((

) )(

( )(

)) )( ))

The results display suitable locations for fern reintroduction, based on the spatial relationship of the weighted criteria. However, the results are categorical. In practical terms, species do not act within the categorical rules, or parameters, defining optimal growth sites. A weighted multi-criteria evaluation assumes that the input data represents reality accurately (GITTA, 2010). Real world applications are not categorical, but rather scaled, and defined by blended ranges. Therefore, it is necessary to re-evaluate the classifications, using a fuzzy overlay analysis. Figure 8, located on page 16 of this document, displays the weighted multi-criteria evaluation results.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 8: MCE Results

GISC9308-D2 February 22, 2013 Lisa Atkinson

6.0 Fuzzy Overlay Analysis Fuzzy overlay analysis is a technique employed to solve model site suitability. This technique is executed using fuzzy logic, an approach which assumes that there exist inaccuracies within the attribute, and geometrically spatial data (Esri, 2012). Therefore, a fuzzy overlay analysis ultimately defines areas of partial memberships, or varying degrees of suitability. 6.1 Evaluation Logic Using Spatial Analyst Fuzzy Membership tool, the hillshade, aspect, slope and soil reclassification results are transformed into fuzzy memberships, to be coalesced with Spatial Analyst Fuzzy Overlay Analysis tool. Required input parameters include membership type, midpoint and spread. The parameters utilized to create the fuzzy membership imagery are defined within Table 3.

Table 3: Fuzzy Membership Parameters

Membership Type Input Data Gaussian Hillshade Reclassified Near Aspect Reclassified Gaussian Slope Reclassified Linear Soil Type - Reclassified

Midpoint 3.0 2.5 3.0 0 (Minimum)

Spread 1.0 0.1 1.0 3 (Maximum)

Gaussian membership, or normal distribution membership, requires the user to enter a midpoint value, defined by maximum fern growth ability, and a spread, defined between 0 and 1, where a larger value indicates a steeper spread around the midpoint (ArcGIS, 2012). Near membership type is similar to Gaussian, in that the user must define the midpoint and range, and generally applicable to compressed data sets (ArcGIS, 2012. Linear membership type defines a fuzzy membership function through a linear transformation, consisting of the minimum and maximum possibility values (ArcGIS, 2012). The Original Terms of Reference document, located in Appendix B, specifies the membership type to be applied to slope, aspect, and hillshade. Figure 9, located on the following page, displays the graphical concept of each membership type.

GISC9308-D2 February 22, 2013 Lisa Atkinson Gaussian Fuzzy Membership Type

Midpoint is defined as the optimal site suitability for fern growth.

Near Fuzzy Membership Type

Midpoint is defined as the middle value within a data set spread.

Linear Fuzzy Membership Type

Figure 9: Membership Types

GISC9308-D2 February 22, 2013 Lisa Atkinson 6.2 Data Transformations The fuzzy membership results, depicted by Figure 10, reflect the varying degrees of suitability on a ramp scale of 0 to 1. A value of 1 indicates full certainty, and a value of 0 indicates unsuitability (Esri, 2012). The parameters outlined by Table 3, are adopted to transform the data. As Figure 10 illustrates, the fuzzy membership results are contained within the range of 0 to 1.

Figure 10: Fuzzy Membership Data Transformations

6.3 Results In order to identify possible suitable sites for fern reintroduction, Spatial Analyst Fuzzy Overlay Analysis tool is employed, coalescing the criteria variables to a site suitability map. Fuzzy overlay analysis incorporates the weighted importance values, previously defined, by setting the Environments Raster Analysis Mask, to the raster calculator generated results. The overlay type is specified as AND, as the input evidence must have a high value, in order for the output value to be high (Esri, 2012). The map: Fuzzy Overlay Analysis Identifying Suitable Planting Areas for Finlays Fabulous Fictitious Fern, located on the following page, denoted as Figure 11, displays the fuzzy overlay analysis results. The legend reflects the ability to identify sites, based on varying degrees of suitability.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 11: Fuzzy Analysis Results

GISC9308-D2 February 22, 2013 Lisa Atkinson 7.0 Recommendations Based on a comparison of the weighted multi-criteria evaluation, and the fuzzy overlay analysis, the recommendation sites to plant, and reintroduce, Finlays Fabulous Fictitious Fern, are indicated by the circled areas, displayed by Figure 12, located on the following page.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 12: Recommended Fern Planting Sites

GISC9308-D2 February 22, 2013 Lisa Atkinson However, due to several existing discrepancies between the two maps, identifying potential suitable sites for fern reintroduction, it is necessary to ground truth the recommended planting sites. For example, Figure 12 depicts a small area, in proximity to the railroad, defined as unsuitable, by the weighted multi-criteria analysis results; but defined as optimal, by the fuzzy overlay analysis results. It is recommended that areas of discrepancies be disregarded, when selecting sites for fern reintroduction. Table 4 compares the pixel count, and therefore the area, identified as suitable, possible, and not recommended, for both spatial analysis techniques.

Table 4: Suitable Area Comparison

Weighted Multicriteria Evaluation Classification Thrive Moderate Growth Sparse Growth Pixel Count 499 6 62

Fuzzy Overlay Analysis Classification Optimal Possible Not Suitable Pixel Count 486 75 6

Table 4 identifies the fuzzy overlay analysis results, incorporating a greater area for possible site selection. However, it is important to analyze both display outputs, as indicated by Figure 12, due to potential discrepancies. In practical terms, species do not act within the categorical rules, or parameters, defining optimal growth sites. However, the fuzzy overlay analysis will typically produce a result with a higher concentration of possible areas. An example of a fuzzy overlay analysis is depicted by Figure 13.

GISC9308-D2 February 22, 2013 Lisa Atkinson

Figure 13: Example of Fuzzy Overlay Analysis

(Source: Esri, 2012)

The results from this investigation do not follow this convention, due to the limited number of categorical ranks utilized. The information is vague and general, producing results that may not best display probable fern reintroduction sites. In this case, it is recommended that the experts be consulted a second time, given a predefined ranking scale, containing 10 characteristic categories, and apply the ranks to the criteria. This may produce a more detailed fuzzy overlay analysis result. Some additional considerations include the following factors, unaccounted for within the criteria listed in the Original Terms of Reference, located in Appendix B of this document: As the fern takes advantage of poor soils, to avoid competition with other species, the possible reintroduction sites must be examined, within the field, to confirm that in fact, no competitive species exist within the planting areas. It may be economically beneficial to begin reintroduction, by planting the fern within a subset of suitable sites, as denoted by Figure 12, and monitor the growth and spread of the fern within the habitat.

GISC9308-D2 February 22, 2013 Lisa Atkinson 8.0 Conclusions The ability of GIS applications, to evaluate multiple criteria, or variables, is economically beneficially. Resources of man power and time, often associated with in field ground truthing, can be substantially reduced. Thus project costs may decrease. It is the requirement of the GIS specialist, to understand the project, in terms of scope, data, and desirable outcomes. This is paramount to identify the correct data manipulation process, in order to produce the most meaningful results. The weighted multi-criteria evaluation analysis, considers each criterion, or variable, as a percentile of importance, in regards to the other criteria. The outcome is typically Boolean, in that results are categorical. However, reality is most often defined in terms of partial memberships, or varying degrees of truth. For this reason, fuzzy membership overlay analysis is becoming increasingly adopted, as the desired data manipulation tool. As discovered in this investigation, there is no perfect spatial analysis result. This reiterates the importance of the GIS specialist to consider all options, and undertake a thorough analysis of the results, to produce meaningful recommendations.

GISC9308-D2 February 22, 2013 Lisa Atkinson 9.0 Bibliography Esri. (2012). ArcGIS 9.2 Desktop Help: About TIN Surfaces. Retrieved February 13, 2013 from http://webhelp.esri.com/arcgisdesktop/9.2/index.cfm?TopicName=About_TIN_surfaces Esri. (2012). ArcGIS Help 10.1: Applying Fuzzy Logic to Overlay Rasters. Retrieved February 13, 2013 from http://resources.arcgis.com/en/help/main/10.1/index.html#//009z000000rv000000 European Commission. (2007). Multi-Criteria Analysis. Retrieved February 13, 2013 from http://forlearn.jrc.ec.europa.eu/guide/4_methodology/meth_multi-criteria-analysis.htm NOAA. (2013). Solar Position Calculator. Retrieved February 4, 2013 from http://www.esrl.noaa.gov/gmd/grad/solcalc/azel.html Raines, G., Sawatzky, D. & Bonham-Carter, G. (2012). Esri: Arc User Online Incorporating Expert Knowledge New Fuzzy Logic Tools in ArcGIS 10.0. Retrieved February 13, 2013 from http://www.esri.com/news/arcuser/0410/fuzzylogic.html

Appendix A
Data Model Diagram

Weighted Multi-Criteria Evaluation Data Model Diagram


Elevation Points Break Lines Study Area Soil Types

3D Analyst Create TIN Tool

Data Management Project Tool Editor Toolbar Soil Location Transformation

TIN

3D Analyst TIN to Raster Tool

Elevation Model

Spatial Analyst Reclassify Tool (With Suitability Field) Spatial Analyst Aspect Surface Tool Spatial Analyst Slope Surface Tool
Escarpment: 3 Toledo/Cashel: 2 Beverley/Peel: 1 Haldimand/Lincoln: 0 Not Mapped: No Data

Spatial Analyst Hillshade Surface Tool


Azimuth: 177.1 Altitude: 65.98

Hillshade

Aspect

Slope

Spatial Analyst Reclassify Tool


0-180: 0 180-195: 1 195-220: 3 220-235: 1 235-255: 0

Spatial Analyst Reclassify Tool


0-247 degrees: 0 247-292.5 degrees: 1 292.5-337.5 degrees: 3 337.5-360 degrees: 1

Spatial Analyst Reclassify Tool


0-20 degrees: 0 20-35 degrees: 1 35-45 degrees: 2 45-55 degrees: 3 55-65 degrees: 1 65-90 degrees: 0

Reclassified Soil Types

Reclassified Hillshade

Reclassified Aspect

Reclassified Slope

Spatial Analyst Raster Calculator Tool


((0.46*Slope)+(0.23*Aspect)+(0.155*Soiltype)+(0.155*Hillshade))* ((Slope/Slope)*(Aspect/Aspect)*(Soiltype/Soiltype)*(Hillshade/Hillshade ))

MCE

Fuzzy Overlay Analysis Data Model Diagram


Reclassified Hillshade Reclassified Aspect Reclassified Slope Reclassified Soil Types

Spatial Analyst Fuzzy Membership


Membership: Gaussian Midpoint: 3 Spread: 1

Spatial Analyst Fuzzy Membership


Membership: Near Midpoint: 2.5 Spread: 0.1

Spatial Analyst Fuzzy Membership


Membership: Gaussian Midpoint: 3 Spread: 1

Spatial Analyst Fuzzy Membership


Membership: Linear Minimum: 1 Maximum: 3

Fuzzy Hillshade

Fuzzy Aspect

Fuzzy Slope

Fuzzy Soil Types

Spatial Analyst Fuzzy Overlay


Environments Raster Analysis Mask: MCE Overlay Type: AND

Fuzzy Overlay Results

You might also like