Advertisement
Not a member of Pastebin yet?
Sign Up,
it unlocks many cool features!
- 1. Determine 4 different spatial weight matrices and calculate the I Moran statistics for variable Donants (Donations to the poor)
- o First-order neighbourhood matrix (queen type)
- weights manager->create->select ID variable->CODE_DE->Create->close-> Moran ScatterPlot->Univariate-> Donants
- o Second-row neighbourhood matrix (without lower orders, queen type)
- weights manager->create->select ID variable->CODE_DE->Order of contiguity-> 2->Create->close-> Moran ScatterPlot->Univariate-> Donants
- o Second-row neighbourhood matrix (with lower orders,(queen type)
- weights manager->create->select ID variable->CODE_DE->Order of contiguity-> 2->include lower orders->Create->close-> Moran ScatterPlot->Univariate-> Donants
- o First-row neighbourhood matrix (rook type)
- weights manager->create->select ID variable->CODE_DE->rook contiguity->Create->close-> Moran ScatterPlot->Univariate-> Donants
- 2. For the matrix of weights with the highest I Moran value:
- o Interpret neighbourhood Moran's statistics
- weak positive spacial autocorrelation
- o Indicate the number of objects located in the HH and LL quadrants
- cluster Maps->univariate Local moran I->Donants->Cluster Map
- 8 i 17
- o How many objects with statistically significant spatial correlation at the level of 0.01 ?
- cluster Maps->univariate Local moran I->Donants->Significance Map
- 3
- 3. Visualise spatial distribution of variable: Donants, using: Quantile map with 6 intervals,percentile map, natural brakes (6) and standard deviation map. What’s the difference between them?
- maps and rates->quantile map->6
- maps and rates->percentile map
- maps and rates->natural breaks map->6
- maps and rates->standard deviation map
- in quantile equally distributed
- in percentile symmetrical with respect to the centre
- in natural breaks intervals are
- 4. Prepare a conditional map for variable: Donants using population size as a condition
- map->conditional map -> Donats, Pop1831, COUNT
- 5. Prepare an anamorphic chart for variable: Donants, use the same variable for circle colour.
- Cartogram -> Donants, Donants
- QGIS
- 1.
- Raster Calculator -> ifelse(a=0,0/0,a/100)
- dla wszystkich
- ifelse(a>25,1,0)*ifelse(b>25,1,0)
- ifelse(gt(abs(a-b),3),0,1)* ifelse(gt(abs(a-c),3),0,1)* ifelse(gt(abs(b-c),3),0,1)
- TABLAU:
- 1. Otwieramy tableau XD
- 2. Te negatywne robimy nowe variable x`=1/x jak x = 0 to x'=c-x (c to average lub mean) [na zajeciach robilismy 1-x] - Aids_New, ChildMortality_New, DrugAddiction_New, HepatitisB_New.
- Dla kazdej zmiennej (aids, childmortality, drugadd, hep) robimy taka, że 1-ona.
- 3. Robimy new calculated field. Dodajemy wszytkie nowe wartosci i Doctors i HospitalBeds - composite indicator. [[Aids_New]+[ChildMortality_New]+[Doctors]+[DrugAdd_New]+[HepatitisB_New]+[Hosptial beds]]
- Dodajemy je plus te pozostale.
- 4. Przeciagamy Voivodship do columns. Calculated indicator do rows.
- 5. Wchodzimy w zakladke Analytics. Przeciagamy Label z napisem Cluster na plansze - tam gdzie sie pojawil wykres. Zamien na kropeczki jak dalej masz te bars dla lamusow. XD
- 6. Liczymy SI - Prawym na Clusters w zakladce Marks -> Describe clusters... SI=(b-q)/(max(a,b)) im blizej 1 tym lepiej. To robimy gdy nie jestesmy pewni czy mamy optymalna ilosc clusters, moze moze byc lepiej! :slight_smile:
- Jak chcemy relacje lekarzy do jakiejs choroby to lecimy tak:
- Przypominam jak chcemy zakolorowac jedna krope to zaznaczamy ja, prawy klik -> group I z prawej strony (mozesz nie widziec jak masz otwarta zakladke Show me) bedziesz widziec to wydzielone cos I "reszte swiata"
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement