PARAMETERS
Times-series parameters
You can slice, transform and otherwise customize your time-series dataset prior to download by appending various optional parameters to your query.
| Parameter | Required | Type | Values | Description |
|---|---|---|---|---|
| database_code | yes | string | Code identifying the database to which the dataset belongs. | |
| dataset_code | yes | string | Code identifying the dataset. | |
| limit | no | int | Use limit=n to get the first n rows of the dataset. Use limit=1 to get just the latest row. | |
| column_index | no | int | Request a specific column. Column 0 is the date column and is always returned. Data begins at column 1. | |
| start_date | no | string | yyyy-mm-dd | Retrieve data rows on and after the specified start date. |
| end_date | no | string | yyyy-mm-dd | Retrieve data rows up to and including the specified end date. |
| order | no | string | asc desc | Return data in ascending or descending order of date. Default is desc. |
| collapse | no | string | none daily weekly monthly quarterly annual | Change the sampling frequency of the returned data. Default is none; i.e., data is returned in its original granularity. |
| transform | no | string | none diff rdiff rdiff_from cumul normalize | Perform elementary calculations on the data prior to downloading. Default is none. Calculation options are described below. |
Collapse: Quandl returns the last observation for any given time period. So if you collapse a daily time-series to monthly, you will get a sample of the original daily time-series where the observation for each month is the last daily data point available for that month. Note that this can cause errors for fields like open price since the daily open price on the last day of the month is not equal to the monthly open price.
Transform: Available transformations are listed in the table below.
Transformations
| Name | Effect | Formula |
|---|---|---|
| none | no effect | z[t] = y[t] |
| diff | row-on-row change | z[t] = y[t] – y[t-1] |
| rdiff | row-on-row % change | z[t] = (y[t] – y[t-1]) / y[t-1] |
| rdiff_from | latest value as % increment | z[t] = (y[latest] – y[t]) / y[t] |
| cumul | cumulative sum | z[t] = y[0] + y[1] + … + y[t] |
| normalize | scale series to start at 100 | z[t] = y[t] ÷ y[0] * 100 |
NOTE:
In the above table, y[0] refers to the starting date specified by
start_dateorlimitand not the starting date of the underlying raw time-series. Similarly,y[latest]refers to the latest date specified byend_date.
Updated over 7 years ago