a) flexibility to choose the start level of granularities - for different applications users might be interested to look at time series behaviour from different start point rather than conventional - 12am to 11pm
b) facets by multiple of primary granularities - for example, every 2 hours, where hour is a primary granularity
c) Instead of facet-ing, just drawing them adjacent to each other, so that sudden jump or fall in other facets clearly visible
d) Looking at the time series of available bikes at any station
e) Comparing time series of rides between station A and station B and also between station B to station A - see if they are related inversely.
f) time series can be looked for (i) trips (ii) refills (iii) stations
g) Think about how package will display missing values. Consider, an irregular time series where there are missing values corresponding to certain time points - can you show them as missing and not a straight line drawn between the non-missing values
Participants - Antony and Sayani
Possible/ Plausible things to work on -
a) flexibility to choose the start level of granularities - for different applications users might be interested to look at time series behaviour from different start point rather than conventional - 12am to 11pm
b) facets by multiple of primary granularities - for example, every 2 hours, where hour is a primary granularity
c) Instead of facet-ing, just drawing them adjacent to each other, so that sudden jump or fall in other facets clearly visible
d) Looking at the time series of available bikes at any station
e) Comparing time series of rides between station A and station B and also between station B to station A - see if they are related inversely.
f) time series can be looked for (i) trips (ii) refills (iii) stations
g) Think about how package will display missing values. Consider, an irregular time series where there are missing values corresponding to certain time points - can you show them as missing and not a straight line drawn between the non-missing values