After long break I decided to come back with new GIS subject. Probably many of you think why I use and somehow promote Quantum GIS in my previous posts while there are lots of other tools available on the software market. Now I had chance to think about the answer. Probably I’m just used to this program and always it offered me exactly what I needed to accomplish my tasks (and its for FREE). It is very good software however I don’t know much about other open-source soultions.
I thought that maybe it is not a bad idea to check other tools in practice!
I searched on the internet topics related to free, open-source GIS software and I found following post:
The article presents the list of 13 free GIS solutions. Let’s try to compare them and see if we agree with the proposed ranking 🙂
I tried to figure out what might be the simplest method of testing the software.
Let’s imagine that we are user who has particular task to do. We need to quickly make a simple map out of available data we have in our inventory. The result of the test is just my personal feeling about the usage of the software. You might have different opinion.
Dataset and testing plan
I’ve put together set of files which will be used by the cartographer and the test will include:
- Website, downloading and installation.
- Loading the data (vector shapefile, georeferenced raster and .csv table).
- Manipulating the data, changing display properties.
- Simple data analysis – interpolation.
- Composing and exporting the map.
Below you can see example dataset: couple of shapefiles, geotiff reaster file, excel file with data. All georeferenced in the same coordinate system (in my case Polish EPSG 2180 coordinate system).
You can find more info on this Coordinate System under this link:
Also the whole list of 13 programs had to be divided into pieces as there were too many to test at once. At the very end maybe I will try to summarize the results for all 13.
For the first group I picked Top 4 from the list: Quantum GIS, gVSIG, Whitebox GAT and SAGA GIS. We will try to use various tools to obtain as much exactly the same map. QGIS will be the reference for me in terms of comparison (that means it will be somehow apart of the test) just because I know it the best.
Downloading and installation
Let’s find our tools on the internet. All of them are very easy to find. Unfortunately some of the websites have not very straightforward addresses such as Whitebox GAT.
Generally all the websites are well designed so the installation files are easy to find by the cartographer under the time pressure. Also there are lots of other information such as project details, reports, guides, tutorials which are worth to read on each of those products.
Both QGIS and gVSIG have modern design, nice looking websites with user friendly download pages. Whitebox GAT site seems to be older design but installation files are also easy to find. SAGA website looks very old-school which might be the promise of very non-user friendly software. Download page might be also little confusing for the new users (it is on Sourceforge).
Installing the tools does not make any problems. All these tools install automatically using installation wizards. The only problem I had were some issues with Java installing gVSIG (you need to make sure if you have the latest Java version).
All the tools present different approach of the design of the user interface. It is hard to judge which soultion is the best however all of them are quite intuitive. Sometimes it is only the matter of training to get use to some of them.
Quantum GIS has classic interface with the main content window, layer tree on the left side and tools on the sides of the screen. After opening gVSIG you might be surprised as you don’t see nothing more that black screen and small menu window with four icons. The window called the Project Manager allows you to add your main map window (called View) which allows you to see the project and its properties and separately window showing data (Tables) and Map which is the map composer. For me this is quite unique solution as I was used to classic project workspace (similar to QGIS). Probably the idea was to easy access to the various projects if you have more than one in your inventory.
Whitebox presents classic approach where you have workspace map view in the main part of the screen and toolbox and layers tree on the sides similarly to QGIS.
SAGA jsut like gVSIG does not open any map window automatically right after start. First you will see empty screen with the toolbox, command line and other windows on the sides. It looks a little scary (but professional) at the first look and gives a feeling of difficult operating.
Setting up the project
We know that our project has to be set up in Polish Coordinate system EPSG 2180. Let’s find out how can we edit project properties.
In QGIS in Project menu you can find and edit all the information about your project such as project title, units of measure and also, what is very important, coordinate system.
You can find there large coordinate systems database and solution which can be useful is to filter the large list using EPSG code to find the proper coordinate system for your project.
gVSIG offers you similar options. Under project properties you can also edit project title, units of measure and coordinate system. It has quite smart CS window which allows you to quickly filter and find the proper coordinate system.
Number three of the ranking – Whitebox GAT surprised me negatively because of the lack of any tools helping to set up project properties. I couldn’t find any menu regarding editing or checking the coordinate system of the project. Probably the coordinate system is just imported together with loaded files but it would be nice to have more control on that feature in GIS software.
In SAGA there is also no general project properties window. However this software offers much more options in this matter in comparison to Whitebox. You can check or change coordinate system in special menu (see slides). I can tell you right now that those menus in SAGA are not very intuitive. You can set up all the parameters of Coordinate System you want to use in this window and it offers you quite a lot of options to edit. It also contains large CS database so you can choose from many coordinate systems. The one disadvantage is that there is no filtering or search options here so you need to know more or less what are you looking for.
Loading the data
In our case we have three kinds of data to load:
- vector shapefiles
- georeferenced raster – as basemap (GEOTIFF)
- excel table with data saved in csv format
1. Loading vector shapefiles
In QGIS you can load your data separately from the menu available on the left side of the layers panel. On the screen you can see with how many formats QGIS can handle. To limit their number and show your files more clearly choose shapefiles. You can load multiple files pressing shift button and selecting a group of files.
After loading you should see your files (two polygon files, point layer and line layer) with the random display properties such as colour and style.
Importing files to gVSIG has totally different procedure. You need to right click on your main map window and select add layers from the menu. Then you need to add you files, select the shapefiles (they should appear on the list) and accept. You should see the layers in the map view.
Whitebox GAT offers very simple data loading method. You just need to pick Data Layers menu and choose Add Layers to Map. Then you can browse for your shapefiles. Interesting thing is that all the file formats available for upload, both raster and vector, are in the same place. After layers were chosen you should see them in your map view and also on the layers tree. Quick and easy.
One of the methods of vector import to SAGA is just to browse for it in the data sources menu (bottom left corner). By doubleclicking of the shapefile it will be added to project database. It is interesting that it will not appear on your map right away after import. It is visible in Data window. This is quite nice tool which seems to be complex at the first sight but it is interesting concept. You can store, review and even see your layers before adding them to the map. If you want to add the data to the map you just need to doubleclick the proper layer in the Data Tree. All the layers can be added to one or more maps. After adding to the same map they should all appear in random style and order. Layer tree is accesible under Maps tab.
What additonal things you can do with your shapefiles? Sometimes you would like to see shapes attributes or measure something. Let’s see how how we can do this in our Top Four tools.
QGIS allows you very easy attributes access with one click. Also you can quickly measure distance, area and angle. Results appear in separate window in selected unit of measure.
In gVSIG you can also very easy access item attributes. Separate window opens to show you the attributes. Also measuring the area and distance is easy. Results are shown on the bottom of the screen.
Whitebox GAT offers nice access to layer attributes. Just one click opens menu on the left side to make the values visible. Distance measurements are easy to make. Result appears on the bottom of the screen. Area measurement is not available from the main menu. You can find it in the toolbox. Description suggests that it might be quite complex and it requires creating of new polygon layer (we don’t have time to do it).
SAGA GIS gives easy access to the layers attributes with one click on the selected feature. They appear in complex looking enviroment but it is only matter of time to get use to it. Distance measurement is easy as this tool is present in the main menu. Result appears in the bottom part of the screen. I couldn’t find Area measurement tool. Even Toolbox offers some complex tool which is not the quick measurement tool which we are looking for in this case.
2. Loading raster files
In our case the raster file is already georeferenced (see also: How to georeference scanned image in QGIS and vectorize items from OpenStreetMap-part 1.) so it should appear in correct position right after import.
In QGIS file import is very similar to the vector files import. Just select Import Raster on the left side menu and browse for your file.
Map should appear on your screen. If it is georeferenced it should already have the project coordinates and will show in a right place, exactly where you expect it to be.
Importing raster in gVSIG also seems to be quite easy. We use the same Add Layer menu in the main map window. After browsing for the file and accepting it in the list it should appear in the map view.
Importing raster in Whitebox GAT turned out to be more complicated. We can import the file using simple Add Layer to Map or using Toolbox. Unfortunately both methods resulted in a strange result in my case. Raster appeared in the right position but it looked weird. I wasn’t able to figure out what happened. If this problem have not appeared importing raster in Whitebox would be very easy.
Importing Raster in SAGA seems also quite easy. You can doubleclick in your data sources similarly to importing shapefiles and then doubleclick it in the Data menu to add it to the map. Also in this case I had problems with the map colors. Map was in the right place but in stange style. Finally I decided to tweak it a little to make it at least black and white changing colors type to Shade (remember after choosing an option you need to press Apply, otherwise change won’t be visible – this is unique thing present only in SAGA).
3. Loading CSV (Excel table)
Our Excel table is simple table with Point locations. It contains only X and Y’s and point ID/name. We will try to import such data to our project.
In QGIS we are using Import CSV button on the left side of the layers tree. We have to select the file and then we should indicate which field of the table shows what parameter (X,Y or name). After that import is finished and points should appear on your map.
In gVSIG theoretically you can also import your CSV table just how we imported shapefile or raster before. However I had problem importing the CSV and after searching for the solution on the various websites I figured out that what is necessary to do is downloading the proper addon called CSV Wizard. After that import went smoothly and points/ Locations appeared on the map.
Importing CSV in Whitebox GAT went surprisingly smooth. You just need to select CSV format from the Add Layer to Map menu.
It is the easiest import but it seems that you don’t have too much control on it. Additionally it converts your CSV file to shapefile so you need to expect new files in your project folder.
SAGA reads your CSV file easily. With doubleclick on your CSV file in the data sources you can import it to the project database and then open it clicking on it. You can see that it was imported but columns and rows were not separated. The walkaround of this issue is that we can import the CSV using toolbox. Import Text Table tool allows us to set up separators. Now based on such table we can convert table to shape using Geoprocessing tab and Convert Table to Points tool. Locations should appear in the database. Doubleclick adds them to the map content. I think that was the most complicated CSV import from all four programs. It gives you more control than Whitebox GAT however it takes time and also it produces additional shapefile which you might not need or want in your folders.
Summary of Part 1.
Both QGIS and gVSIG have the most intuitive and the widest project properties setup with large Coordinate Systems databases. SAGA is more difficult to operate; it also offers you large coordinate systems database, but not so many properties to edit. Whitebox is very poor in this matter not having similar project properties setup tool. Importing data is quite easy and quick in all of the tools. Interesting concept is to import data without differentiation wheater those are raster or vector files – this concept is used in Whitebox or gVSIG. Importing raster was problematic both for SAGA and Whitebox GAT from unknown reasons. Both QGIS and gVSIG coped with this task very well. Importing CSV created in MS Excel was challenging for SAGA, it required small processing of the data. On the other hand gVSIG required installation of new add-on especially for CSV import. Both finally ended with success but I had to use Google to find those solutions. Importing file in Whitebox was the easiest and fastest however it gives you the impression that you don’t have full control on your data.
This post got very long so the comparison of top 4 GIS open-source tools from http://gisgeography.com will be continued in the next part. We will edit imported layers and finally produce a map!